TerraClear's mission to help farmers map and tackle tough field problems such as rocks and weeds has evolved with the launch of a new machine: an autonomous robot called TerraScout. “TerraScout will scout entire fields in almost any condition and turn that intelligence into precise action for existing crews and equipment,” TerraClear CEO Devin Lammers said in a new release Thursday. “Today we focus that output on rock and weed management, but the future applications for this platform are vast. It is my firm belief that this technology will drive the next era of farm productivity gains.” TerraScout can operate autonomously for up to six hours without refueling. Onboard technology turns massive image datasets into actionable maps in real-time. Field trials for TerraScout began earlier this year and TerraClear said it will expand trials to existing retail partners and select farm customers this spring. Tech Moves: TerraClear's new CEO; CoreWeave hires ex-Oracle SVP; Amperity adds revenue chief Carbon Robotics raises $20M as LaserWeeder maker plans secretive new ‘AI robot' for farms Fertile ground for innovation: A list of agriculture tech startups in Seattle and Washington state WSU researchers develop an AI robot that uses a puff of air to find and pick ripe strawberries

The U.S. State Department is reportedly working on an online portal that would allow people in Europe and other regions to access content banned by their governments. The move comes at a time when conservative figures like Elon Musk and J.D. Vance have railed against European attempts to clamp down on hate speech, terrorist propaganda, and revenge porn. Reuters reported Wednesday, citing unnamed sources, that the initiative is intended to fight censorship and could include a virtual private network (VPN) feature. The site currently displays a landing page featuring a small animation of Paul Revere on horseback above the words “Freedom is Coming.” Smaller text below reads, “Information is power. “Free speech, I fear, is in retreat and in the interests of comedy, my friends, but also in the interest of truth, I will admit that sometimes the loudest voices for censorship have come not from within Europe, but from within my own country, where the prior administration threatened and bullied social media companies to censor so-called misinformation,” Vice President J.D. Vance said in his speech last week at the Munich Security Conference. The State Department did not immediately respond to a request for comment from Gizmodo. However, the agency told Reuters it does not have a Europe-specific censorship bypass program and denied that any announcement had been postponed. “Digital freedom is a priority for the State Department, however, and that includes the proliferation of privacy and censorship-circumvention technologies like VPNs,” a State Department spokesperson told Reuters. The portal also comes as Elon Musk's social media platform X is facing mounting scrutiny in Europe. Musk has been an on-and-off ally to President Donald Trump. Paris prosecutors' cybercrime unit, working alongside Europol and French national police, raided X's offices in the country earlier this month. The investigation began in January 2025 over suspected manipulation of X's recommendation algorithm and illegal data extraction. It has since expanded to include complicity in the spread of pornographic images involving minors, the use of sexual deepfakes that infringe on people's rights, and the circulation of Holocaust denial content. The European Commission has also slapped X with a $140 million fine for violating transparency obligations under the Digital Services Act, taking issue with the platform's “deceptive design” of its blue checkmark verification badges and a lack of transparency around advertising practices. “Rumors swirling that the EU commission will fine X hundreds of millions of dollars for not engaging in censorship. The EU should be supporting free speech not attacking American companies over garbage,” Vance wrote on X before the fine was announced. Reuters also reports that the portal project involves Edward Coristine, a former member of the Musk-led Department of Government Efficiency (DOGE). Coristine now works with the National Design Studio, created under Trump to redesign federal services online and in physical spaces. The studio has contributed to several Trump-era initiatives, including Trump Accounts and TrumpRx. The initiative funded grassroots tools to help people bypass government internet controls worldwide. Donald Trump Jr. insists his father is "not involved."

At the very least, it's a place where ideas are being hatched by tech founders who are inspired by living and working with one another. Bili House is a hacker house started by a group of young people interested in improving connections and opportunities in the Seattle-area startup community. And applications are open for a first cohort of four to six teams. The house was launched by four founders: Sylviane Zhao, who recently graduated from Cornell University, and Shawn Yang and Tehani Cabour, who both worked at French software giant Dassault Systèmes. They're working together on projects including CodeChimp, a project management platform that aims to turn vibe coding into a “multiplayer experience” by using multi-agent orchestration and other AI-powered tools. Last fall they were part of a Plug and Play cohort in Seattle. Jatin Kumar is the fourth founding member and a Z Fellow. “We're just trying to get the early stage startup scene kick started here in Bellevue,” Yang told GeekWire. “Every morning you wake up, you just go upstairs from your room and start working with each other,” Zhao added. Yang said that before starting the hacker house, they were considering a move to San Francisco. That really changed my perspective,” Yang said, adding that he feels like there are more startup “doers” than just “talkers” choosing to live and work this way. Tech veteran Andy Rebele (Pure Watercraft) ran a few different spaces more than a decade ago, including on Capitol Hill and in the University District. Seattle startup Tune also ran a house in 2015 near the University of Washington for women studying computer science. The desire for houses geared specifically toward female entrepreneurs continues today with FoundHer House, a San Francisco-based space spotlighted by The New York Times last year. The Bili House website says rent ranges from $500 to $2,000 per month depending on room size. Amenities include utilities, high-speed internet, access to all common spaces, and community events. In addition to events such as demo nights, founder dinners, and hackathons, the group is looking into partnerships, perhaps with a venture capital firm that could help defer some costs for startup founders. Other AI startups currently working out of Bili House include legal simulation platform LexSims and construction cost analysis company Bevr. Keep scrolling for more images of Bili House: Have a scoop that you'd like GeekWire to cover? San Francisco had an all-female AI ‘Hacker House' — is Seattle ready for its own? Plug and Play's second Seattle-area accelerator cohort includes six local startups Startup Radar: Seattle founders build AI tools for leadership training, spatial reasoning, vibe coding New startup hubs supercharge Seattle's tech scene — just in time for the AI boom

SoftBank subsidiary SB Energy is expected to build a massive 9.2 gigawatt natural gas-fired power plant on the Ohio-Kentucky border, according to a report from Bloomberg. With a price tag of $33 billion, the project would be more expensive than recent natural gas-fired power plants, which have skyrocketed in cost, Bloomberg notes. It's unclear who will ultimately foot the bill, though traditionally rate-payers have shouldered the burden for new generating capacity. SB Energy did not say whether the new power plant would feed directly into the grid or if it would power data centers. OpenAI and SoftBank are in the process of building a “proof of concept” data center at GM's former Lordstown automotive assembly plant. A power plant of this size is likely to take years, perhaps a decade, to complete even before taking into account the shortage of natural gas turbines. If completed, the project could emit around 15 million metric tons of carbon dioxide per year, according to our calculations based on public, energy-use metrics. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. Every weekday and Sunday, you can get the best of TechCrunch's coverage. TechCrunch Mobility is your destination for transportation news and insight. Startups are the core of TechCrunch, so get our best coverage delivered weekly. Provides movers and shakers with the info they need to start their day. By submitting your email, you agree to our Terms and Privacy Notice.

On February 19, 2025, a Falcon 9 rocket fell back toward Earth in an uncontrolled reentry, producing a massive fireball in the skies over Europe. On its way down, the rocket came near a lidar station in Saxony, Germany, where a team of researchers was able to use the remote-sensing instrument to measure the effect of the rocket reentry on Earth's atmosphere. As spacecraft reenter through the atmosphere, they slowly disintegrate into smaller fragments that burn up and release metals into the upper atmosphere. The researchers behind the new study chose to measure lithium because it's an effective tracer of space debris pollution and is commonly used in spacecraft. “We discussed several materials, and lithium gave a good compromise of expected signal and challenges in the lidar setup,” Wing said. Although the risks of space debris falling onto the ground are well-studied, there is very little known about the effects of falling spacecraft on Earth's upper atmosphere. Using lidar—a laser-based remote sensing instrument used to measure atmospheric conditions—the team of scientists behind the new study detected a sudden increase in the concentration of lithium atoms, around 10 times the baseline value found in the atmosphere, on February 20, 2025. The recent study focuses on the amount of lithium left behind from a single rocket reentry. Previous studies have already shown that lithium, aluminum, copper, and lead left behind from the reentry of spacecraft already exceeds the cosmic dust influx of those metals to the atmosphere. It's not clear yet how exactly this change in the atmospheric chemistry would affect life on Earth, but the researchers behind the new study believe that there is enough reason for concern. Still, more research is needed to narrow down the impact on our planet. “There are many elements present inside spacecraft which are not very present in our atmosphere due to natural causes,” Wing said. “We know very little about what metals actually exist in the atmosphere and how that relates to re-entry pollution.” Subscribe and interact with our community, get up to date with our customised Newsletters and much more. Local environmentalists warn that the island's marine ecosystem could be placed at great risk. “There is still a great deal of work ahead to prepare for this historic mission.” Just days after SpaceX pivoted to the Moon, Blue Origin's new lunar strategy has come to light.

OpenAI is nearing a deal to raise more than $100 billion at a valuation that could exceed $850 billion, Bloomberg reports, citing sources familiar with the matter. To that end, OpenAI has said it has started testing ads in ChatGPT for free users, a gamble that could lead to more revenue or could send users running from the platform. The company's pre-money value will remain at $730 billion, per Bloomberg's source. The first tranches of funding are reportedly coming from the usual suspects: Amazon (already in talks to invest up to $50 billion), SoftBank (gearing up for $30 billion), Nvidia (close to investing $20 billion), and Microsoft. VC firms and sovereign wealth funds are expected to close later, potentially bringing the total amount raised higher. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. Every weekday and Sunday, you can get the best of TechCrunch's coverage. TechCrunch Mobility is your destination for transportation news and insight. Startups are the core of TechCrunch, so get our best coverage delivered weekly. Provides movers and shakers with the info they need to start their day. By submitting your email, you agree to our Terms and Privacy Notice.

The global rise in nearsightedness has largely been attributed to our phone addiction, but researchers now suggest that incessant scrolling is not solely to blame. A team of scientists from the SUNY College of Optometry found that dim indoor lighting, combined with prolonged, up-close focus, may be straining the eyes. Shortsighted people have slightly stretched eyeballs, making it difficult to focus on distant objects. Although the increasing number of shortsighted people has largely been linked to excessive use of our phones, the new study suggests it's not that simple. Instead, focusing on objects up close in low-light environments strains the retina. “When people focus on close objects indoors, such as phones, tablets, or books, the pupil can also constrict, not because of brightness, but to sharpen the image. In dim lighting, this combination may significantly reduce retinal illumination.” Prolonged focus on objects in dim lighting further enhances this pupil constriction, especially for people who are already nearsighted. When the light is too dim, it causes the pupils to narrow excessively at short viewing distances, making it difficult to develop normal vision. The theory is speculative and requires further testing to confirm. “But the study offers a testable hypothesis that reframes how visual habits, lighting, and eye focusing interact.” The condition described here is myopia, which requires glasses for distance. Good news: There are other things to worry about. California scientists have potentially created a non-invasive method of reshaping the cornea. In a new case report, doctors in France detail how the woman developed a rare form of retina damage tied to hair dyes made with aromatic amines. Signify's light and door monitoring functions are fully Matter compatible, though the new Philips Hue Secure cameras are still waiting for Matter support.

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. In a respite from the ever-increasing cost of PC hardware and accessories, Dell has released a brand new gaming monitor for less than $130. At this price point, outside of a sale, you can typically expect low refresh rates and aging hardware. This Dell SE2726HG monitor bucks the trend – it's a 27-inch monitor with an ultra-fast 240Hz refresh rate, and all for an introductory sale price of just $129.99 right now. The stand-out feature on this display is the 240Hz refresh rate, which is rarely seen at this price, and leaves plenty of room to keep up with your GPU's output in any fast-paced games you might play. It ships with a 27-inch IPS panel, a good all-round choice on LCD monitors these days, as it typically offers the best viewing angles. Other features on this display include two HDMI 2.1 ports, along with a single DisplayPort 1.4 connection. The display uses a tilt-capable mount, so you do have flexibility to adjust its position, but if you want a height-adjustable stand included, you can with the otherwise-identical Dell SE2726HGS, albeit at the higher price of $169.99. The budget-friendly SE2726HG, meanwhile, has a max brightness of 300 nits. It also has a standard response time of 5 ms, but this can (reportedly) be reduced to just 0.5ms by switching to the display's ‘extreme' mode in the settings. Nvidia G-Sync support isn't mentioned, but you'll often find the two work interchangeably when one or the other is supported on a display, although we can't confirm that here. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. Ben Stockton is a deals writer at Tom's Hardware. He's been writing about technology since 2018, with bylines at PCGamesN, How-To Geek, and Tom's Guide, among others. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

The primary purpose of CPF is not a pension scheme. It is structured as a massive forced bond purchase scheme by citizens. Financially what happens is the 37% of citizen income buys a long term bond (till retirement age, on average decades) at rock bottom interest rates (it's pegged to the overnight rate or a minimum of 2.6%). This has historical roots in the government needing vast capital financing. They make enormous amounts of the delta between the short term interest rate and long term capital gains. Singapore has no oil or natural resources, but it's sovereign wealth fund has AUM in the regions of countries like Norway which do for this reason. It is not a shock absorber like the article suggests. The withdrawal terms are strict - housing, a significant medical expense and retirement are the only real ways to get money out of it. In fact the Singapore government maintains a large worker supply through immigration. After all this there's the astronomical cost of living. This is also intentional, to raise the number of employees. In fact the Singapore government maintains a large worker supply through immigration. After all this there's the astronomical cost of living. This is also intentional, to raise the number of employees. The reason it raising the retirement age is effective in workforce participation is because most people have no choice. After all this there's the astronomical cost of living. This is also intentional, to raise the number of employees. It's essentially a forced loan to the government at subpar rates. Most other countries have lower rates on their retirement schemes, which makes it feasible for more people to live on their savings for a few years before the government retirement scheme kicks in. E.g. in the US it's pretty feasible for the upper middle/lower upper classes to retire a few years before Social Security kicks in, especially if they're willing to live frugally. Most other countries have lower rates on their retirement schemes, which makes it feasible for more people to live on their savings for a few years before the government retirement scheme kicks in. E.g. in the US it's pretty feasible for the upper middle/lower upper classes to retire a few years before Social Security kicks in, especially if they're willing to live frugally. But you have to hope that Republicans - and yes this is a political issue - don't successfully kill the ACA and make it impossible to get insurance at any cost if you have a pre-existing condition. But my mom was a teacher who still gets benefits through the government and my dad gets benefits from the one factory that didn't shut down in our hometown.I'm 51 and even if I could retire early financially, I wouldn't do it and stay in the US. I “retired my wife” at 44 in 2020 8 years into our marriage when I did a slight transition to an industry where remote work with travel is the norm (cloud consulting + app dev) and we have traveled a lot including doing stints as “digital nomads”.We are staying in one of the countries that we might retire to as a Plan B for six weeks starting next week.Even now that we moved to state tax free Florida and my wife hasn't had to work in six years, she keeps a current CDL because she can get a job as a school bus driver easily for the benefits and someone will pay me for independent consulting if I lose my job. But my mom was a teacher who still gets benefits through the government and my dad gets benefits from the one factory that didn't shut down in our hometown.I'm 51 and even if I could retire early financially, I wouldn't do it and stay in the US. I “retired my wife” at 44 in 2020 8 years into our marriage when I did a slight transition to an industry where remote work with travel is the norm (cloud consulting + app dev) and we have traveled a lot including doing stints as “digital nomads”.We are staying in one of the countries that we might retire to as a Plan B for six weeks starting next week.Even now that we moved to state tax free Florida and my wife hasn't had to work in six years, she keeps a current CDL because she can get a job as a school bus driver easily for the benefits and someone will pay me for independent consulting if I lose my job. I “retired my wife” at 44 in 2020 8 years into our marriage when I did a slight transition to an industry where remote work with travel is the norm (cloud consulting + app dev) and we have traveled a lot including doing stints as “digital nomads”.We are staying in one of the countries that we might retire to as a Plan B for six weeks starting next week.Even now that we moved to state tax free Florida and my wife hasn't had to work in six years, she keeps a current CDL because she can get a job as a school bus driver easily for the benefits and someone will pay me for independent consulting if I lose my job. We are staying in one of the countries that we might retire to as a Plan B for six weeks starting next week.Even now that we moved to state tax free Florida and my wife hasn't had to work in six years, she keeps a current CDL because she can get a job as a school bus driver easily for the benefits and someone will pay me for independent consulting if I lose my job. Even now that we moved to state tax free Florida and my wife hasn't had to work in six years, she keeps a current CDL because she can get a job as a school bus driver easily for the benefits and someone will pay me for independent consulting if I lose my job. All incompatible with 99% of the upper class, neither do they want to eat ramen to retire early.You're also one medical disaster away from being "very very wrong" I bet you also your idea of upper middle class is not statistically valid.https://dqydj.com/household-income-percentile-calculator/ So the top 10% is a household income of $250K and most of those couples didn't reach that until their 40s. They aren't making $225K as an L5 at 25 years old like a former intern/new grad I mentored when I was at BigTechMost software developers won't even see above $160K inflation adjusted during their career. Most software developers won't even see above $160K inflation adjusted during their career. You can shop around quite a bit for non urgent care, and get good cash discount. Yeah there's even a term for it: https://en.wikipedia.org/wiki/Financial_repression It also robs the individual's freedom to gamble with their retirement funds while expecting/demanding a bailout when shit hits the fan.In the USA we have thoughtful policies that allow people over a certain amount of wealth invested in key industries to do that. In the USA we have thoughtful policies that allow people over a certain amount of wealth invested in key industries to do that. Edit: in fact interest delta is how banks make their huge profits except the government here does it by force. It is a tax, but with extra steps.The reason it makes the government money is because they're collecting the extra interest that citizens would have earned if they were free to invest it on their own. The reason it makes the government money is because they're collecting the extra interest that citizens would have earned if they were free to invest it on their own. As for why - the same reason why they get to decide what side of the road you drive on and what laws you follow. They rule the patch of land you were born on, and if you don't like it you can either participate in the system (assuming it's a democracy) or leave. "There is no moral justification for the government setting a retirement age, but they are able to. Then steal whatever they have left with medical bills and price hikes on necessities. Hell, social security was initially resented by older workers because of the cover it gave employers for firing them for being too old. Social Security constitutionality was ruled on just months after the 'switch in time that saved 9' associated with a threatening to pack the courts and evade the checks and balances built into our "democracy." They ruled it was covered under 'general welfare' in a way that was totally historically inaccurate.Furthermore, FDR and congress purposefully had it packaged in an omnibus style bill to evade democratic scrutiny over the individual portions, by purposefully torpedoing other aid to needy individuals if SS didn't pass, so that lawmakers wouldn't be able to vote on democratic view of SS but rather being damned in a catch-22 where they'd be accused of not helping out the needy in other ways.Basically the whole thing was designed to not only evade democracy but also the constitution. Furthermore, FDR and congress purposefully had it packaged in an omnibus style bill to evade democratic scrutiny over the individual portions, by purposefully torpedoing other aid to needy individuals if SS didn't pass, so that lawmakers wouldn't be able to vote on democratic view of SS but rather being damned in a catch-22 where they'd be accused of not helping out the needy in other ways.Basically the whole thing was designed to not only evade democracy but also the constitution. To make an overly dramatic analogy, if you were kidnapped and asked why the kidnapper was able to hold you against your will, the answer is because they've chained you up and they have the gun, and so on. The government makes a moral justification of a savings plan but then when we dig down to it it's all ether and really just a scheme for bond rate arbitrage for the government.The point isn't that might makes right is false, it's that the moral justification is a facade. You might as well ask similar questions about most basic laws and concepts behind how western societies work. There's maybe a few hundred people worldwide who could casually drop a proper answer to your question while casually browsing hn.I believe it'd be more fair to start answering your own question to show how far you are in your intellectual journey on that topic. I believe it'd be more fair to start answering your own question to show how far you are in your intellectual journey on that topic. And the reason it decides that, apart from "because it can", is because many societies have seen what happens when it's left to individuals to take care of this, and they fuck it up in massive numbers, and the outcome of that then fucks up society. It sounds to me like we have built a system to exploit people as much as possible. > Your duty as a citizen is to work and build your nationWhat about the duty of the trust fund babies and idle wealthy? Why is the retirement age going up instead of down as productivity increases? Why is the retirement age going up instead of down as productivity increases? In the case of 401(k)s/DC plans and private pensions/DB plans, the government allowed savings without "confiscation," i.e. immediate taxation. They gave us the benefit of deferred taxation if you wait until retirement age. The current trajectory makes my question a lot of things, including this whole "government pays back that service with benefits" as it will be some time before I ever see a penny of SSI.A lot of our taxes in this country seem like a giant waste or are grossly inefficient at best. A lot of our taxes in this country seem like a giant waste or are grossly inefficient at best. Our government is filled with grifters because we've failed to hold them meaningfully accountable for robbing us and failing to provide the benefits we're funding.Many of the grifters in government have been working hard to make it difficult to hold them accountable. They disenfranchise voters, they keep us afraid and our futures uncertain, they collude against efforts to reform the system they've established for their own benefit.Government was never going to just let us have "liberty and justice for all" the job was always on "we the people" to insist on it. We can't just pay taxes and expect everything to work out. If we've reached a point where that's no longer possible then it's our duty to "refresh the tree of liberty" until we have a government that works for us. Many of the grifters in government have been working hard to make it difficult to hold them accountable. They disenfranchise voters, they keep us afraid and our futures uncertain, they collude against efforts to reform the system they've established for their own benefit.Government was never going to just let us have "liberty and justice for all" the job was always on "we the people" to insist on it. We can't just pay taxes and expect everything to work out. If we've reached a point where that's no longer possible then it's our duty to "refresh the tree of liberty" until we have a government that works for us. Government was never going to just let us have "liberty and justice for all" the job was always on "we the people" to insist on it. We can't just pay taxes and expect everything to work out. If we've reached a point where that's no longer possible then it's our duty to "refresh the tree of liberty" until we have a government that works for us. It's essentially no different than payroll taxes in the U.S. that fund Social Security. Buying government bonds is still marginally better accounting than a complete Ponzi scam like Social Security in the U.S., but even that ultimately amounts to the same thing - the government is paying itself, so it's a wash. The only question is whether the fund is running at a surplus or not.The US has raided its fund to finance other government programs, and then will have to pay it back via tax revenues. The US has raided its fund to finance other government programs, and then will have to pay it back via tax revenues. The parent contributor has conveniently left out the fact that the 37% of CPF contributions is split 20-17 in terms of employee-employer contributions[1], and has a ceiling of S$8000[2], so if one earns more than that, every additional dollar goes entirely to them, which is also taxed at globally low income tax rates[3]. This point is a shell game, because the employer's share is still effectively being taken from the employee. It's equivalent of "tariffs are paid by foreigners!" Forcing your own citizens to save money for themselves is a lot better than forcing your own citizens to pay for others. Forcing your own citizens to save money for themselves is a lot better than forcing your own citizens to pay for others. HK has a similar forced savings, but that ROI is like 1 or 2% and the options to invest are paltry.Some perspective is necessary. Forced investment in low ROI vehicles is just a tax by another name. Do people actually get kicked out at 99 years, or does the government renew it for a nominal fee? It's not as if the US fairs any better. Eminent domain means that they can take your property from you at any time even if you've kept paying them.Singapore has a home ownership rate of ~90% vs 65% in the US and prices are so unaffordable that on average people buying their first starter homes in the US are in their 40s! Most Americans, if they're lucky will get maybe get 30 years of their life in a home they own while they are still young and healthy enough to enjoy it.Last I heard Singapore only had about 1,000 homeless. Whatever they're doing with housing could probably be improved on, but they seem to be doing a lot better than we are. Singapore has a home ownership rate of ~90% vs 65% in the US and prices are so unaffordable that on average people buying their first starter homes in the US are in their 40s! Most Americans, if they're lucky will get maybe get 30 years of their life in a home they own while they are still young and healthy enough to enjoy it.Last I heard Singapore only had about 1,000 homeless. Whatever they're doing with housing could probably be improved on, but they seem to be doing a lot better than we are. Last I heard Singapore only had about 1,000 homeless. Whatever they're doing with housing could probably be improved on, but they seem to be doing a lot better than we are. In Singapore, the government owns everything, even ostensibly 'freehold' land. Eminent domain exists in the U.S. and other developed countries too. The point of it is largely to prevent any single owner from "holding up" a non-trivial project like an MRT line. I grew up in East Germany, and while it was a total failure, they got at least one idea correct in the workers paradise: We need to work. (Never mind the implementation, I already said it was a total failure, okay? I'm actually like my grandfather, who without any need whatsoever continued to work well past retirement, privately, painting a house here, doing some paint shop there, designing and installing a sun dial somewhere. He only got off the scaffolding on a house's paint job a week before he died.I too would hate to just laze around. I worked and made money many times as a child already, and it was always fun!What stopped the fun was the coming of The West (which I too went to the streets for and wanted, still, "side effects may apply"). Being in the production of stuff is actually FUN! Except then came some western management idiot to make it clear fun is over. Just a few minutes, no time to start something else. In comes the management idiot, immediately jumping on me, why am I lazing around??? That's not what they pay me for!Just an anecdote, and of course it is much better in knowledge jobs, but that, and the fact that the money accumulates towards the top is what I think is a HUGE problem in today's capitalism. No wonder they have to make live as miserable as possible for the working majority, because there is no fun. But it is THEM who are responsible for much of that. I'm actually like my grandfather, who without any need whatsoever continued to work well past retirement, privately, painting a house here, doing some paint shop there, designing and installing a sun dial somewhere. He only got off the scaffolding on a house's paint job a week before he died.I too would hate to just laze around. I worked and made money many times as a child already, and it was always fun!What stopped the fun was the coming of The West (which I too went to the streets for and wanted, still, "side effects may apply"). Being in the production of stuff is actually FUN! Except then came some western management idiot to make it clear fun is over. Just a few minutes, no time to start something else. In comes the management idiot, immediately jumping on me, why am I lazing around??? That's not what they pay me for!Just an anecdote, and of course it is much better in knowledge jobs, but that, and the fact that the money accumulates towards the top is what I think is a HUGE problem in today's capitalism. No wonder they have to make live as miserable as possible for the working majority, because there is no fun. But it is THEM who are responsible for much of that. I too would hate to just laze around. I worked and made money many times as a child already, and it was always fun!What stopped the fun was the coming of The West (which I too went to the streets for and wanted, still, "side effects may apply"). Being in the production of stuff is actually FUN! Except then came some western management idiot to make it clear fun is over. Just a few minutes, no time to start something else. In comes the management idiot, immediately jumping on me, why am I lazing around??? That's not what they pay me for!Just an anecdote, and of course it is much better in knowledge jobs, but that, and the fact that the money accumulates towards the top is what I think is a HUGE problem in today's capitalism. No wonder they have to make live as miserable as possible for the working majority, because there is no fun. But it is THEM who are responsible for much of that. Being in the production of stuff is actually FUN! Except then came some western management idiot to make it clear fun is over. Just a few minutes, no time to start something else. In comes the management idiot, immediately jumping on me, why am I lazing around??? That's not what they pay me for!Just an anecdote, and of course it is much better in knowledge jobs, but that, and the fact that the money accumulates towards the top is what I think is a HUGE problem in today's capitalism. No wonder they have to make live as miserable as possible for the working majority, because there is no fun. But it is THEM who are responsible for much of that. Just an anecdote, and of course it is much better in knowledge jobs, but that, and the fact that the money accumulates towards the top is what I think is a HUGE problem in today's capitalism. No wonder they have to make live as miserable as possible for the working majority, because there is no fun. But it is THEM who are responsible for much of that. Payroll taxes are collected and placed in the social security trust fund, which invests them in federal bonds. Currently there are more payouts than taxes so the trust fund is being used to make up the difference.When the trust fund is depleted (barring any changes, this happens at some point in the next decade if I'm not mistaken) then there will be a reckoning. When the trust fund is depleted (barring any changes, this happens at some point in the next decade if I'm not mistaken) then there will be a reckoning. The Obama administration explicitly appropriated general government funds to try and make up a developing shortfall in the 'fund'. There is no money being accumulated because there are more payouts than taxes - but even if that wasn't the case, these are not actual "bonds" that have been bought on any market, they're just non-market government obligations. 100% respect for them.A side comment: I enjoy listening to English language news from many countries around the world to get different viewpoints. News media from Singapore is very interesting, indeed! A side comment: I enjoy listening to English language news from many countries around the world to get different viewpoints. News media from Singapore is very interesting, indeed! We are talking about material impacts, not culture People who score well on probability numeracy are likely better educated and better paid and have more in automatic savings plans. So if someone is maxing out their 401k they don't feel they need to save more.The article shows that in the US there is a 25 point gap between high and low income on savings regret, and a 14 point gap between high and low numeracy scores.In Singapore where savings are more automatic numeracy is a more powerful predictor. The article shows that in the US there is a 25 point gap between high and low income on savings regret, and a 14 point gap between high and low numeracy scores.In Singapore where savings are more automatic numeracy is a more powerful predictor. Considering that they also have to consider economic development in their investment decisions, the RRQ funds are well managed by the CDPQ. If you can structure the above as insurance - and you can trust the insurance - (I know a few cases where the insurance type system went bankrupt and those with a "policy got nothing") that is best.Once the above is taken care of though, you can't take it with you (at least in most religions) so spend it. Once the above is taken care of though, you can't take it with you (at least in most religions) so spend it. You're looking for people who didn't want to save but begrudgingly saved at the expense of their pre-retirement life and then died before they could enjoy retirement. But in general you have three things to spend in your life: time, money, and effort.You don't want to spend all your time saving money because you'll run out of time eventually, but you also don't want to spend all your money saving time because you'll run out of money.It's all about balance and thoughtfully understanding what you actually want and how to get it. You don't want to spend all your time saving money because you'll run out of time eventually, but you also don't want to spend all your money saving time because you'll run out of money.It's all about balance and thoughtfully understanding what you actually want and how to get it. This becomes more true by the year, as those costs keep rising faster than broader inflation. I get it to a certain extent, don't live in poverty if you don't have too, but I am a major saver. I rarely buy new things if an old thing is working fine. If I die early at least my family will will be set.Really the social safety nets in the US are basically non-existent so having a big savings buffer makes me feel a bit safer. Honestly dying early doesn't worry me too much, I'll be dead so doesn't bother me. What does worry me is the economy tanks and all my saving become worthless. Really the social safety nets in the US are basically non-existent so having a big savings buffer makes me feel a bit safer. Honestly dying early doesn't worry me too much, I'll be dead so doesn't bother me. What does worry me is the economy tanks and all my saving become worthless. Yes the jobloss impact caused the people to be unable to save and in turn they wished they have saved more.. but ignored is whether they could to begin with.Of course external impact had little to do with internal procrastination. Of course external impact had little to do with internal procrastination. It says that understanding risk (as operationalized by understanding probability) has a larger effect.But it is also saying that the more external impact someone has, the more they regret saving more -- in the United States but not Singapore.The study is explicitly saying that internal motivation does not seem to matter. But it is also saying that the more external impact someone has, the more they regret saving more -- in the United States but not Singapore.The study is explicitly saying that internal motivation does not seem to matter. The study is explicitly saying that internal motivation does not seem to matter. It's basically just saying that the uninsured catastrophic event risk in America magnifies shock events.E.g., if you have a major hospital visit in America you're way more likely to regret not saving enough, but in Singapore there's basically no effect since hospital stays don't drain your savings account. E.g., if you have a major hospital visit in America you're way more likely to regret not saving enough, but in Singapore there's basically no effect since hospital stays don't drain your savings account. Not just healthcare stuff, but also apparently Singaporeans tend to have a lower unemployment rate, so they be able to recover from stuff faster ------ re: below due to throttling -------------Vs say US, where immigrants and those funding public housing are generally better off than the people getting subsidized housing. Public housing is more a progressive than regressive tax in the US, so quite dissimilar. Asking "but how is this any different" (after I already answered it, lol) over and over doesn't negate this, nor the fact that immigrants are like half of workers in Singapore vs only 10% in the US so the funding dynamic and dependency is far different. Vs say US, where immigrants and those funding public housing are generally better off than the people getting subsidized housing. Public housing is more a progressive than regressive tax in the US, so quite dissimilar. Asking "but how is this any different" (after I already answered it, lol) over and over doesn't negate this, nor the fact that immigrants are like half of workers in Singapore vs only 10% in the US so the funding dynamic and dependency is far different. It's similar in Vienna where only native Viennese are immediately eligible for social housing, but outsiders will end up paying into the system without being eligible. By definition, outsiders don't have to pay into the system since they already have a gov't somewhere else that is dedicated to them, just like the Viennese do. By outsiders in this context I meant foreign workers. Countries like Singapore and all of the Middle East meanwhile rely on a revolving door of cheap immigrant labor. Most of these countries don't have a pathway to citizenship at all for this worker class. It doesn't matter how many numbers you put in spreadsheets dated for 20-40 years into the future, if in said future, there isn't actually anyone to accept those numbers in exchange for labor. That seems like what they should have been looking at re procrastination--conscientiousness.I am not at all surprised that people who Take Care of Business lament not doing a better job (saving) and people who YOLO don't as much. I am not at all surprised that people who Take Care of Business lament not doing a better job (saving) and people who YOLO don't as much. It turned on hardcore capitalism mode for itself because a significant portion of its population wants to try and solo socioeconomic hardships and hates any one who doesn't want the same challenge.But not to just blame the voter, lots of money is spent for setting up systems to be amenable to acquiring more money. The very richest have correctly made a bet that uprisings to displace the wealthy and politicians just don't occur here these days, and therefore there is no real threat or need to change the way things have been going for the last 25 years or so. The very richest have correctly made a bet that uprisings to displace the wealthy and politicians just don't occur here these days, and therefore there is no real threat or need to change the way things have been going for the last 25 years or so. Yes, the systems are amenable to acquiring more money, but I would claim that all that the richest need to do is to push the idea that "anyone can make it" - which was probably (more) true 50 years ago, but is probably an illusion today (some comments at: https://en.wikipedia.org/wiki/Socioeconomic_mobility_in_the_...).Edit: I do not claim one model is better than the other; just that the culture influences the outcome more than other aspects.

Seattle startup Adronite raised $5 million in a Series A round led by Gatemore Capital Management, as it looks to expand its AI-powered platform designed to give large organizations visibility into sprawling and complex codebases. Unlike many AI coding tools that operate at the level of individual files or snippets, Adronite ingests complete codebases, including both modern and legacy systems. The idea is to help organizations understand how their software works as a system, with applications in security analysis, modernization, and active remediation at scale. Adronite can also build apps from natural language prompts and offers an AI chat feature that provides details on system-wide insights. The system supports more than 20 programming languages and has been proven on a codebase with 2.5 million lines of code. There are various companies that offer code review tools, including CAST and Sonar. Adronite says its software differentiates in part due to its private network settings, as it can run entirely inside a secured internal environment. Have a scoop that you'd like GeekWire to cover? Seattle's DeltaGen raises $1.2M to market its generative AI tech for streamlining M&A deals Seattle startup Gradial raises $35M to boost agentic tools that automate enterprise marketing Seattle startup takes a swing at enterprise search with new AI-fueled discovery tool

Sam Basu quit his job as a senior software engineer at Google in early 2023, not long after OpenAI released ChatGPT. Basu got “very curious” and started cold-calling customs brokers in the Los Angeles area. He learned that many are mom-and-pop affairs still deeply reliant on fax machines and paper. He flew to that customer's office the next day. This seed of an idea is now called Amari AI, a startup co-founded by Basu and Arushi Vashist, a former senior software engineer at LinkedIn. The pair, and their small team, have already collected more than 30 customers, and helped those firms move more than $15 billion of goods. Amari has also raised $4.5 million of funding co-led by preeminent early-stage firms First Round Capital and Pear VC, all before coming out of stealth mode on Thursday. Some leverage optical character recognition software to help speed up data entry, but that tech is limited and brittle, he said. Amari is meant to help automate data entry and paperwork, and let employees — who legally must be in the United States, meaning companies can't use off-shore workers — focus on helping customers move their goods across the border. President Donald Trump's chaotic trade policy has made customs brokers all the more important, according to Chris Bachinski, CEO of 125-year-old firm GHY International. Bachinski — who is one of Amari's early adopters — told TechCrunch that many of his own customers don't even have their own compliance staff. With the employee base tightly regulated, and a licensing exam pass rate that hovers around 10% to 20%, “it's a perfect fit for AI,” he said. “Experienced people are leaving the industry or taking early retirement,” Basu said. “So we're pitching ourselves as these extra set of hands that logistics companies can hire or keep alongside the human expertise.” Basu said Amari's AI agents constantly monitor trade rules and update their reasoning any time there's a change, making it easier for brokers to help their customers quickly understand any impact. Previously, these kinds of sudden changes required manual research that would slow down brokers' ability to clear cargo into the country. Amari does this by building its own AI models trained on more than one million documents related to the shipments it's already helped clear through customs, though Basu said the company has been leveraging off-the-shelf models to date. GHY International is not a mom-and-pop, but it's also not a Fortune 500 like FedEx; Bachinski has been looking for ways to both stay competitive and grow his company. Bachinski said the biggest concern amongst GHY employees so far has been job loss. He said he expects tech like Amari's to help GHY grow, and to focus more on the customer relationships and compliance work. And with trade now constantly in the spotlight, he said brokers need to be agile. “I make this joke that last year, the first time in history, our families know what we do for a living. Most recently, he was a reporter at Bloomberg News where he helped break stories about some of the most notorious EV SPAC flops. Hear from 250+ tech leaders, dive into 200+ sessions, and explore 300+ startups building what's next. Meta's own research found parental supervision doesn't really help curb teens' compulsive social media use How Ricursive Intelligence raised $335M at a $4B valuation in 4 months Hollywood isn't happy about the new Seedance 2.0 video generator The great computer science exodus (and where students are going instead) A Stanford grad student created an algorithm to help his classmates find love; now, Date Drop is the basis of his new startup

When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. “Teleporting quantum information is now a practical reality,” asserts Deutsche Telekom. The firm's T‑Labs used commercially available Qunnect hardware to demo quantum teleportation over 30km of live, commercial Berlin fiber, running alongside classical internet traffic. In an email to Tom's Hardware, Deutsche Telekom's PR folks said that Cisco also ran the same hardware and demo process to connect data centers in NYC. Last month, T-Labs completed the first practical test of the core components required for a future quantum internet, which would work by teleporting data. Central to the demo was Qunnect's Carina platform, which integrates an entanglement generator, producing pairs of quantum-entangled photons for distribution over telecom fiber. “Our fiber optic network is quantum ready,” said Abdu Mudesir, Telekom Board Member for Product and Technology. “In Berlin we have now proven that quantum information can be transmitted over 30 kilometers of commercial Telekom fiberoptics outside of a laboratory. There are a few wrinkles still to iron out, though. Still, it is pleasing that this milestone has been reached, with the building blocks of teleportation already operating across a real network. The teleportation wavelength used was 795nm, which is said to be a sweet spot for integration with platforms such as neutral-atom quantum computers, atomic clocks, and various quantum sensors. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. This milestone achievement shifts quantum communications from an experimental lab technology towards something telecoms providers can deploy. We last reported on Cisco's quantum internet efforts back in November last year, when it announced plans to jointly build a distributed quantum computing network capable of linking fault-tolerant systems over long distances, with the help of IBM. Deutsche Telekom will also have a ‘Quantum Teleportation' showcase at its booth. Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds. Mark Tyson is a news editor at Tom's Hardware. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

OpenAI looks for more funding as it aims to fulfill spending goals When you purchase through links on our site, we may earn an affiliate commission. Get Tom's Hardware's best news and in-depth reviews, straight to your inbox. OpenAI may be on the cusp of securing a new stockpile of cash to burn through, with the business looking to secure $100 billion in funding from a range of tech and investment firms, Bloomberg reports. With $1.4 trillion in pledged expenditure over the next eight years, OpenAI has enormous commitments to meet. While OpenAI is also preparing for an IPO later this year, its plans have not gone without criticism, with some reportedly coming from Nvidia CEO Jensen Huang himself. Nvidia invested in companies that often bought Nvidia chips, while tech firms investing in AI developers often sold those same companies cloud computing capacity. The key players in the deal are Amazon, Nvidia, Microsoft, and SoftBank, all of which have a mix of deals with one another and OpenAI already. Amazon's investment is rumored to be up to $50 billion, with the condition of that investment being that OpenAI uses more of its chips and cloud computing services. Although Microsoft's name was also thrown around by Bloomberg's sources, there was no speculation on what kind of financial contribution it would make to this round of funding. All companies involved are expected to finalize their investments by the end of February. As with other major investment deals, even the rumored news of this one sent stock prices rising. AI industry sceptics would suggest this is one more example of why the AI industry is in a bubble: Investments that haven't happened yet generate financial returns for these companies, even without any clear path to profitability. Microsoft CEO Satya Nadella was seeming wobbled by the "slop" moniker, and bubble name-calling; Nvidia's Jensen Huang specifying the details of its $100 billion partnership with OpenAI, and an increasing number of studies suggesting AI isn't helping improve productivity much, if at all. OpenAI is even projected to run out of cash entirely by 2027. OpenAI also released GPT 5.3 Codex, which is the first of OpenAI's deployments that marks a move away from being reliant solely upon Nvidia chips. Once the major tech firms have signed on, OpenAI is reportedly looking to generate further investment from venture capital firms, sovereign wealth funds, and other financial investors, leading to its eventual IPO. Although it's not explicitly stated, the fact that OpenAI is targeting these other sources after its industry contemporaries have taken the first step feels rather deliberate. With significant spending commitments already booked in, OpenAI is clearly looking toward how it might fulfill the titanic amount of spending. At the AI India summit, it's still making more commitments for new data center deals, expending more money it doesn't yet have, on the back of confidence generated by deals that have yet to be finalized, or in some cases even officially announced. Jon Martindale is a contributing writer for Tom's Hardware. Tom's Hardware is part of Future US Inc, an international media group and leading digital publisher.

Epstein tragically was committed suicide, and no other cases are forthcoming. Any commoner would have been sent "quick smart" to face the accusations there in court? (I'm kind of amazed he chose that name, tbh; it's not particularly uncommon for British monarchs to rename themselves on taking the throne, and it has... Example - Why all the supposed "...rich and powerful names ...." being seemingly protected ?What do they have to hide ? * When did other high profile people know about this illegal and immoral behavior* Who else is getting away with similar behavior right now * Who else is getting away with similar behavior right now There's no way there isn't at least one other person that's started to fill the void.Ideally, the ramifications of association with Epstein should shrink the size of the vacuum considerably, but the pursuit of those associates has really only just started and, as someone else has already pointed out, some countries / governments are protecting these associates rather than investigating / prosecuting. Also, it might be an anomaly that one person has a very big network ; maybe it's usually more of a "small adjacent networks".So it would be like asking "who replaced pablo escobar or bernard madoff". So it would be like asking "who replaced pablo escobar or bernard madoff". 2 - Isn't it convenient that zero major news organizations - controlled by high profile people and their buddies - are raising that issue? Not that I believe there to be any public support for competent & systematic enforcement of the laws against such behavior. That I've heard of, nobody even cares about how Jeff got off with a slap on the wrist in 2008. (Sadly, expecting the Yanks to follow their lead on that would be pure fantasy.)

Claude Code is a lock in, where Anthropic takes all the value.If the frontend and API are decoupled, they are one benchmark away from losing half their users.Some other motivations: they want to capture the value. Even if it's unprofitable they can expect it to become vastly profitable as inference cost drops, efficiency improves, competitors die out etc. Or worst case build the dominant brand then reduce the quotas.Then there's brand - when people talk about OpenCode they will occasionally specify "OpenCode (with Claude)" but frequently won't.Then platform - at any point they can push any other service.Look at the Apple comparison. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. If the frontend and API are decoupled, they are one benchmark away from losing half their users.Some other motivations: they want to capture the value. Even if it's unprofitable they can expect it to become vastly profitable as inference cost drops, efficiency improves, competitors die out etc. Or worst case build the dominant brand then reduce the quotas.Then there's brand - when people talk about OpenCode they will occasionally specify "OpenCode (with Claude)" but frequently won't.Then platform - at any point they can push any other service.Look at the Apple comparison. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. Some other motivations: they want to capture the value. Even if it's unprofitable they can expect it to become vastly profitable as inference cost drops, efficiency improves, competitors die out etc. Or worst case build the dominant brand then reduce the quotas.Then there's brand - when people talk about OpenCode they will occasionally specify "OpenCode (with Claude)" but frequently won't.Then platform - at any point they can push any other service.Look at the Apple comparison. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. Then there's brand - when people talk about OpenCode they will occasionally specify "OpenCode (with Claude)" but frequently won't.Then platform - at any point they can push any other service.Look at the Apple comparison. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. But Apple also gets to charge Google $billions for being the default search engine. They get to sell cloud storage, and even somehow a TV. That's all super profitable.At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. At some point Claude Code will become an ecosystem with preferred cloud and database vendors, observability, code review agents, etc. This coding agent is minimal, and it completely changed how I used models and Claude's cli now feels like extremely slow bloat.I'd not be surprised if you're right in that this is companies / management will prefer to "pay for a complete package" approach for a long while, but power-users should not care for the model providers.I have like 100 lines of code to get me a tmux controls & semaphore_wait extension in the pi harness. That gave me a better orchestration scheme a month ago when I adopted it, than Claude has right now.As far as I can tell, the more you try to train your model on your harness, the worse they get. I'd not be surprised if you're right in that this is companies / management will prefer to "pay for a complete package" approach for a long while, but power-users should not care for the model providers.I have like 100 lines of code to get me a tmux controls & semaphore_wait extension in the pi harness. That gave me a better orchestration scheme a month ago when I adopted it, than Claude has right now.As far as I can tell, the more you try to train your model on your harness, the worse they get. I have like 100 lines of code to get me a tmux controls & semaphore_wait extension in the pi harness. That gave me a better orchestration scheme a month ago when I adopted it, than Claude has right now.As far as I can tell, the more you try to train your model on your harness, the worse they get. As far as I can tell, the more you try to train your model on your harness, the worse they get. The reason these LLM tools being good is they can "just do stuff." I'll just have my other tool use Claude Code in tmux. If third party agents can be banned from doing stuff (some advanced always on spyware or whatever), then a large chunk of the promise of AI is dead.Amp just announced today they are dumping IDE integration. Models seem to run better on bare-bones software like Pi, and you can add or remove stuff on the fly because the whole things open source. Someone like Microsoft can also provide a SLA and price such appropriately. If you don't own a business or funding source, the way state of the art LLMs are being used today is totally uneconomical (easy $200+ an hour at API prices. )36+ months out, if they overbuild the data centers and the revenue doesn't come in like OpenAI & Anthropic are forecasting, there will be a glut of hardware. If that's the case I'd expect local model usage will scale up too and it will get more difficult for enterprise providers. (Nothing is certain but some things have become a bit more obvious than they were 6 months ago.) Amp just announced today they are dumping IDE integration. Models seem to run better on bare-bones software like Pi, and you can add or remove stuff on the fly because the whole things open source. Someone like Microsoft can also provide a SLA and price such appropriately. If you don't own a business or funding source, the way state of the art LLMs are being used today is totally uneconomical (easy $200+ an hour at API prices. )36+ months out, if they overbuild the data centers and the revenue doesn't come in like OpenAI & Anthropic are forecasting, there will be a glut of hardware. If that's the case I'd expect local model usage will scale up too and it will get more difficult for enterprise providers. (Nothing is certain but some things have become a bit more obvious than they were 6 months ago.) Someone like Microsoft can also provide a SLA and price such appropriately. If you don't own a business or funding source, the way state of the art LLMs are being used today is totally uneconomical (easy $200+ an hour at API prices. )36+ months out, if they overbuild the data centers and the revenue doesn't come in like OpenAI & Anthropic are forecasting, there will be a glut of hardware. If that's the case I'd expect local model usage will scale up too and it will get more difficult for enterprise providers. (Nothing is certain but some things have become a bit more obvious than they were 6 months ago.) 36+ months out, if they overbuild the data centers and the revenue doesn't come in like OpenAI & Anthropic are forecasting, there will be a glut of hardware. If that's the case I'd expect local model usage will scale up too and it will get more difficult for enterprise providers. (Nothing is certain but some things have become a bit more obvious than they were 6 months ago.) (Nothing is certain but some things have become a bit more obvious than they were 6 months ago.) If I'm in a competitive industry that slow down alone can mean failure. Is there any conceivable GUI overlay that Anthropic or OpenAI can add to make their software better than the current terminal apps? Sure, for certain edge cases, but then why isn't the user building those themselves? 24 months ago we could have said that's too hard, but that isn't the case in 2026.Microsoft added all of this stuff in to Windows, and it's a 5 alarm fire. Stuff that used to be usable is a mess and really slow. Running linux with Claude Code, Codex, or Pi is clearly superior to having a Windows device with neither (if it wasn't possible to run these in Windows; just a hypothetical. Monday morning, an the Anthropic API endpoint is down, uh oh! 2030, the data center expansion is bust, nVidia starts selling all of these cards to consumers directly and has a huge financial incentive to make sure the performant local models exist. Everyone in the semiconductor supply chain below nvidia only cares about keeping sales going, so it stops with them.Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? Microsoft added all of this stuff in to Windows, and it's a 5 alarm fire. Stuff that used to be usable is a mess and really slow. Running linux with Claude Code, Codex, or Pi is clearly superior to having a Windows device with neither (if it wasn't possible to run these in Windows; just a hypothetical. Monday morning, an the Anthropic API endpoint is down, uh oh! 2030, the data center expansion is bust, nVidia starts selling all of these cards to consumers directly and has a huge financial incentive to make sure the performant local models exist. Everyone in the semiconductor supply chain below nvidia only cares about keeping sales going, so it stops with them.Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? Monday morning, an the Anthropic API endpoint is down, uh oh! 2030, the data center expansion is bust, nVidia starts selling all of these cards to consumers directly and has a huge financial incentive to make sure the performant local models exist. Everyone in the semiconductor supply chain below nvidia only cares about keeping sales going, so it stops with them.Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? If the end game is just the same as talking to the Star Trek computer, and competitors are narrowing gaps rather than widening them (e.g. Anthropic and OpenAI releases models minutes from each other now, Chinese frontier models getting closer in capability not further), then it is really hard to see how either company achieves a vertical lock down.We could actually move down the stack, and then the real problem for OpenAI and Anthropic is nVidia. 2030, the data center expansion is bust, nVidia starts selling all of these cards to consumers directly and has a huge financial incentive to make sure the performant local models exist. Everyone in the semiconductor supply chain below nvidia only cares about keeping sales going, so it stops with them.Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? 2030, the data center expansion is bust, nVidia starts selling all of these cards to consumers directly and has a huge financial incentive to make sure the performant local models exist. Everyone in the semiconductor supply chain below nvidia only cares about keeping sales going, so it stops with them.Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? Maybe nvidia is the real winner?Also is it just me or does it now feel like hn comments are just talking to a future LLM? Also is it just me or does it now feel like hn comments are just talking to a future LLM? The advances in the Claude Code harness have been more around workflow automation rather than capability improvements, and truthfully workflows are very user-dependent, so an opinionated harness is only ever going to be "right" for a narrow segment of users, and it's going to annoy a lot of others. There are parallels to the silly Metaverse hype wave from a few years ago. At the time I saw a surprising number of people defending the investment saying it was important for Facebook to control their own platform. Well sure it's beneficial for Facebook to control a platform, but that benefit is purely for the company and if anything it would harm current and future users. This whole game is a bizarre battle.In the future, many companies will have slightly different secret RL sauces. I'd want to use Gemini for documentation, Claude for design, Codex for planning, yada yada ... there will be no generalist take-all model, I just don't believe RL scaling works like that.I'm not convinced that a single company can own the best performing model in all categories, I'm not even sure the economics make it feasible.Good for us, of course. In the future, many companies will have slightly different secret RL sauces. I'd want to use Gemini for documentation, Claude for design, Codex for planning, yada yada ... there will be no generalist take-all model, I just don't believe RL scaling works like that.I'm not convinced that a single company can own the best performing model in all categories, I'm not even sure the economics make it feasible.Good for us, of course. I'm not convinced that a single company can own the best performing model in all categories, I'm not even sure the economics make it feasible.Good for us, of course. With how comically extensible pi is and how much control it gives you over every aspect of the pipeline, as soon as you start building extensions for your own, personal workflow, Claude Code legimitely feels like a trash app in comparison.I don't care what Anthropic does - I'll keep using pi. Allow other to take all the value you provide? Asking people who pay for a personal subscription rather than paying by the API call to use that subscription themselves sounds to me like it is. Rather like the way I pay for my internet connectivity with a fixed monthly bill. Sure, someone with extremely heavy usage might not be able to use a normal consumer internet subscription; but it works fine for my personal use. Rather like the way I pay for my internet connectivity with a fixed monthly bill. Sure, someone with extremely heavy usage might not be able to use a normal consumer internet subscription; but it works fine for my personal use. This statement is plainly wrong.If you boost and praise AI usage, you have to face the real cost.Can't have your cake and eat it, too. If you boost and praise AI usage, you have to face the real cost.Can't have your cake and eat it, too. Can't have your cake and eat it, too. If they can provide a relatively cheap subscription against the direct API use, this is because they can control the stuff end-to-end, the application running on your system (Claude Code, Claude Desktop) and their systems.As you subscribe to these plans, this is the "contract", you can use only through their tools. If you want full freedom, use the API, with a per token pricing.For me, this is fair. If you want full freedom, use the API, with a per token pricing.For me, this is fair. Their costs are not magically lower when you use claude code vs when you use a third-party client.> For me, this is fair.This is, plain and simple, a tie-in sale of claude code. I am particularly amused by people accepting it as "fair" because in Brazil this is an illegal practice. I am particularly amused by people accepting it as "fair" because in Brazil this is an illegal practice. This is, plain and simple, a tie-in sale of claude code. I am particularly amused by people accepting it as "fair" because in Brazil this is an illegal practice. I am very curious what is particularly illegal about this. You are paying for access to Claude Code.Is it also illegal that if you pay for Playstation Plus that you can't play those games on an Xbox?Is it illegal that you can't use third party netflix apps?I really don't want to defend and AI company here but this is perfectly normal. In no other situation would we expect access to the API, the only reason this is considered different is because they also have a different service that gives access to the API. Now we all know obviously the API is being used because that is how things work, but you are not actually paying a subscription for the API. You are paying for access to Claude Code.Is it also illegal that if you pay for Playstation Plus that you can't play those games on an Xbox?Is it illegal that you can't use third party netflix apps?I really don't want to defend and AI company here but this is perfectly normal. In no other situation would we expect access to the API, the only reason this is considered different is because they also have a different service that gives access to the API. Is it also illegal that if you pay for Playstation Plus that you can't play those games on an Xbox?Is it illegal that you can't use third party netflix apps?I really don't want to defend and AI company here but this is perfectly normal. In no other situation would we expect access to the API, the only reason this is considered different is because they also have a different service that gives access to the API. Is it illegal that you can't use third party netflix apps?I really don't want to defend and AI company here but this is perfectly normal. In no other situation would we expect access to the API, the only reason this is considered different is because they also have a different service that gives access to the API. I really don't want to defend and AI company here but this is perfectly normal. In no other situation would we expect access to the API, the only reason this is considered different is because they also have a different service that gives access to the API. Anthropic provides an API third-party clients can use. While I do personally disagree with thinking that you should be able to do this when it was never sold in that way, at the end of the day as a customer you can choose if you want to use the product in the way that they are saying or use something else if you don't want to support that model.However the person I was responding too brought up legality which is a very different discussion. However the person I was responding too brought up legality which is a very different discussion. But if that is the service they are making and they are clear about what it is when you sign up... That does not make it illegal.I can see why people think they should be entitled to do this, but it does not align with how they are selling the service or how many other companies sell services. In most situations you don't get unlimited access to the individual components of how a service works (the API), you are expected to use the service (in this case Claude Code) directly. I can see why people think they should be entitled to do this, but it does not align with how they are selling the service or how many other companies sell services. In most situations you don't get unlimited access to the individual components of how a service works (the API), you are expected to use the service (in this case Claude Code) directly. "Both parties are okay with the terms" is far from being sufficient to make something "legal".Tie-in sales between software and services is not different from price dumping. Tie-in sales between software and services is not different from price dumping. MMO's (and gaming in general) being a major example of this, but so are many of the apps I pay for subscriptions for on my phone.The actual technical implementation of how it works is irrelevant when it is clear what it is you are paying for.> "Both parties are okay with the terms" is far from being sufficient to make something "legal".True but the opposite is also true, just because you don't like the terms it does not make it illegal. MMO's (and gaming in general) being a major example of this, but so are many of the apps I pay for subscriptions for on my phone.The actual technical implementation of how it works is irrelevant when it is clear what it is you are paying for.> "Both parties are okay with the terms" is far from being sufficient to make something "legal".True but the opposite is also true, just because you don't like the terms it does not make it illegal. The actual technical implementation of how it works is irrelevant when it is clear what it is you are paying for.> "Both parties are okay with the terms" is far from being sufficient to make something "legal".True but the opposite is also true, just because you don't like the terms it does not make it illegal. > "Both parties are okay with the terms" is far from being sufficient to make something "legal".True but the opposite is also true, just because you don't like the terms it does not make it illegal. True but the opposite is also true, just because you don't like the terms it does not make it illegal. And in many cases like Claude Code and the Anthropic models, they can and do work perfectly independently.> True but the opposite is also true, just because you don't like the terms it does not make it illegal.This is not me "not liking it". This practice is clearly not done to favor the consumer. This practice is clearly not done to favor the consumer. This practice is clearly not done to favor the consumer. Is Spotify, Apple Music, etc etc etc also illegal in Brazil?It would be ridiculous to argue that I could pay for a subscription to World of Warcraft and make my own third party client to play the game with. (Obviously you are free to argue it all you want but I would be very surprised if this was actually illegal).> And in many cases like Claude Code and the Anthropic models, they can and do work perfectly independently.Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?As far as the Anthropic models, yes like many other services they ALSO have a public API that is separate from the subscription that you are paying for.The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. (Obviously you are free to argue it all you want but I would be very surprised if this was actually illegal).> And in many cases like Claude Code and the Anthropic models, they can and do work perfectly independently.Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?As far as the Anthropic models, yes like many other services they ALSO have a public API that is separate from the subscription that you are paying for.The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. > And in many cases like Claude Code and the Anthropic models, they can and do work perfectly independently.Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?As far as the Anthropic models, yes like many other services they ALSO have a public API that is separate from the subscription that you are paying for.The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?As far as the Anthropic models, yes like many other services they ALSO have a public API that is separate from the subscription that you are paying for.The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. As far as the Anthropic models, yes like many other services they ALSO have a public API that is separate from the subscription that you are paying for.The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. The critical difference here being that in the subscription it is very clear that you are paying for “Claude Code” which is a combination of an application and a server side component. It makes no claims about API usage as part of your subscription, again the technical implementation of the service you are actually paying for “Claude Code” is irrelevant.When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for, they could be sending the information to Gemini or or a human looks at it. "Tie-in sale": the business practice where a seller conditions the sale of one product (the tying good) on the buyer's agreement to purchase a different product (the tied good).The examples you are giving are not "tie-in" sales because the service from Playstation Plus, Spotify, Apple Music, etc is the distribution of digital goods.> Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?Which part are you not understanding?I don't care about Claude Code. All I care about is the access to the models through the client that I was already using!> When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! The examples you are giving are not "tie-in" sales because the service from Playstation Plus, Spotify, Apple Music, etc is the distribution of digital goods.> Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?Which part are you not understanding?I don't care about Claude Code. All I care about is the access to the models through the client that I was already using!> When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! > Unless I am mistaken Claude Code does not have a local model built into it, so it requires a server side component to work?Which part are you not understanding?I don't care about Claude Code. All I care about is the access to the models through the client that I was already using!> When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! Which part are you not understanding?I don't care about Claude Code. All I care about is the access to the models through the client that I was already using!> When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! All I care about is the access to the models through the client that I was already using!> When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! > When it comes to “Claude Code” for all that we should care about, again given that “Claude Code” is what you are paying for.No, it is not! If that was true, then getting equivalent usage of the API without claude.ai and Claude Code should cost less, not more.You can try to find all sorts of explanations for it, at the end of the day is quite simple: they are subsidizing one product in order to grow the market share, and they are doing it at a loss now, because they believe they will make up for it later. I understand the reasoning from a business point of view, but this doesn't mean they are entitled to their profits. I do not understand people that think we simply accept their premise and assume they can screw us over just because they asked and put it on a piece of paper. You can try to find all sorts of explanations for it, at the end of the day is quite simple: they are subsidizing one product in order to grow the market share, and they are doing it at a loss now, because they believe they will make up for it later. I understand the reasoning from a business point of view, but this doesn't mean they are entitled to their profits. I do not understand people that think we simply accept their premise and assume they can screw us over just because they asked and put it on a piece of paper. - OpenCode did not break any security protocol in order to integrate with them. - OAuth is *precisely* a system to let third-party applications use their resources. The fact that I was a customer does not mean that I need to protective of their profits.> (from their business perspective)So what? !Basically, they set up an strategy they thought it was going to work in their favor (offer a subsidized service to try to lock in customers), someone else found a way to turn things around and you believe that we should be okay with this? !Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? - OpenCode did not break any security protocol in order to integrate with them. - OAuth is *precisely* a system to let third-party applications use their resources. The fact that I was a customer does not mean that I need to protective of their profits.> (from their business perspective)So what? !Basically, they set up an strategy they thought it was going to work in their favor (offer a subsidized service to try to lock in customers), someone else found a way to turn things around and you believe that we should be okay with this? !Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? !Basically, they set up an strategy they thought it was going to work in their favor (offer a subsidized service to try to lock in customers), someone else found a way to turn things around and you believe that we should be okay with this? !Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? !Basically, they set up an strategy they thought it was going to work in their favor (offer a subsidized service to try to lock in customers), someone else found a way to turn things around and you believe that we should be okay with this? !Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? Basically, they set up an strategy they thought it was going to work in their favor (offer a subsidized service to try to lock in customers), someone else found a way to turn things around and you believe that we should be okay with this? !Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? Honestly, I do not understand why so many people here think it is fine to let these huge corporations run the same exploitation playbook over and over again. Basically they set up a mouse trap full of cheese and now that the mice found a way to enjoy the cheese without getting their necks broken, they are crying about it? If something isn't explicitly provided in the contract, then it can be changed at any point in any way without notice.Honestly, I'm not big on capitalism in general, but I don't understand why people should expect companies to provide things exactly the way they want at exactly the prices they would like to be charged (if at all). That's just not how the world/system works, or should, especially given there are so many alternatives available. If one doesn't like what's happening with some service, then let the wallet do the talking and move to another. Emigration is a far more effective message than complaining. Honestly, I'm not big on capitalism in general, but I don't understand why people should expect companies to provide things exactly the way they want at exactly the prices they would like to be charged (if at all). That's just not how the world/system works, or should, especially given there are so many alternatives available. If one doesn't like what's happening with some service, then let the wallet do the talking and move to another. Emigration is a far more effective message than complaining. Those wanting direct access to the internal service can get an API key for that purpose. That's just how their product offering is structured. What I don't understand is why start this game of cat and mouse? YT-DLP, and the dozens of apps that use it, basically use Youtube's unofficial web API and it still works even after Youtube constantly patches their end. Though now, YT-DLP has to use a makeshift JS interpreter and maybe even spawn Chromium down the line. So if they can get away with routing 80% of the requests to Haiku and only route to Opus for the requests that really need it, that does give them a cost model where they can rely on lower costs than if a third-party client just routes to Opus for everything. They still have the total consumption under their control (*bar prompt caching and other specific optimizations) where in the past they even had different quotas per model, it shouldn't cost them more money, just be a worse/different service I guess As things are currently, better models mean bigger models that take more storage+RAM+CPU, or just spend more time processing a request. All this translates to higher costs, and may be mitigated by particular configs triggered by knowledge that a given client, providing particular guarantees, is on the other side. If subsidizing that offering is a good hook to get higher paying API users on board, then some of that cost is a customer aquisition cost, whereas the cost to them of providing the API doesn't have the same proportion that they can justify as a customer acquisition cost. Netflix: limits number of devices and stream quality and offline use.AWS: does not allow any number of applications (spamming, crypto mining, adult content)Airlines: do not allow smoking, boom boxesIs there any service that gives complete freedom? AWS: does not allow any number of applications (spamming, crypto mining, adult content)Airlines: do not allow smoking, boom boxesIs there any service that gives complete freedom? Airlines: do not allow smoking, boom boxesIs there any service that gives complete freedom? Is there any service that gives complete freedom? There's this pervasive idea left over from the pre-llm days that compute is free. This sounds like engineering, finance, and legal got together and decided they were in an untenable position if OpenAI started nudging OpenClaw to burn even more tokens on Anthropic (or just never optimize) + continually updated workarounds to using subscription auth. But I'm sure OpenAI would never do something like that...At the end of the day, it's the same 'fixed price plan for variable use on a constrained resource' cellular problem: profitability becomes directly linked to actual average usage. Not possible: OpenClaw is run by a foundation, and is open source, which means OpenAI has no leverage to do such a thing. You can of course use OpenCode or any other project with the API, which is also offered as a separate product. You are paying for usage via their application.If their business plan is based on how quickly a human can enter requests and react to the results, and Claude Code is optimized for that, why should you be allowed to use an alternative client that e.g. always tries to saturate the token limits? If their business plan is based on how quickly a human can enter requests and react to the results, and Claude Code is optimized for that, why should you be allowed to use an alternative client that e.g. always tries to saturate the token limits? The rate limits aren't everything.If agent harnesses other than Claude Code consume more tokens than average, or rather, if users of agent harnesses other than CC consume more tokens than average, well, Anthropic wouldn't be unhappy if those consumers had to pay more for their tokens. If agent harnesses other than Claude Code consume more tokens than average, or rather, if users of agent harnesses other than CC consume more tokens than average, well, Anthropic wouldn't be unhappy if those consumers had to pay more for their tokens. Probably the ToS change was to make it more clear.To be fair, the developer is the one breaking the ToS in the most significant way, breaking boilerplate reverse engineering clauses.But the user also is very aware that they are doing something funny, in order to authenticate, the user is asked to authorize Claude Code, n ot Opencode or OpenClaw, it's clearly a hack and there is no authorization from Anthropic to OpenClaw, and you are not giving Anthropic authorization to give access to OC, the user asks Anthropic to give access to Claude Code, the only reason this works is because OC is pretending to be Claude Code.The bottom line issue is that as a user you are paying for a subscription to a package that includes an expected usage. That package is not metered, but it is given on the condition that you will use it as it is expected to be used, by chatting manually with the chatbot, which results in a reasonable expected token usage. By using a program that programatically calls the chat interface, the token consumption increases beyond what was part of the original deal, and so the price should be different.A similar scenario would be if you go to an all you can eat buffet, you pay for a single person, but then you actually unload an army of little clones that start eating the whole buffet. To be fair, the developer is the one breaking the ToS in the most significant way, breaking boilerplate reverse engineering clauses.But the user also is very aware that they are doing something funny, in order to authenticate, the user is asked to authorize Claude Code, n ot Opencode or OpenClaw, it's clearly a hack and there is no authorization from Anthropic to OpenClaw, and you are not giving Anthropic authorization to give access to OC, the user asks Anthropic to give access to Claude Code, the only reason this works is because OC is pretending to be Claude Code.The bottom line issue is that as a user you are paying for a subscription to a package that includes an expected usage. That package is not metered, but it is given on the condition that you will use it as it is expected to be used, by chatting manually with the chatbot, which results in a reasonable expected token usage. By using a program that programatically calls the chat interface, the token consumption increases beyond what was part of the original deal, and so the price should be different.A similar scenario would be if you go to an all you can eat buffet, you pay for a single person, but then you actually unload an army of little clones that start eating the whole buffet. That package is not metered, but it is given on the condition that you will use it as it is expected to be used, by chatting manually with the chatbot, which results in a reasonable expected token usage. By using a program that programatically calls the chat interface, the token consumption increases beyond what was part of the original deal, and so the price should be different.A similar scenario would be if you go to an all you can eat buffet, you pay for a single person, but then you actually unload an army of little clones that start eating the whole buffet. That package is not metered, but it is given on the condition that you will use it as it is expected to be used, by chatting manually with the chatbot, which results in a reasonable expected token usage. By using a program that programatically calls the chat interface, the token consumption increases beyond what was part of the original deal, and so the price should be different.A similar scenario would be if you go to an all you can eat buffet, you pay for a single person, but then you actually unload an army of little clones that start eating the whole buffet. It's just price differentiation - they know consumers are price sensitive, and that companies wanting to use their APIs to build products so they can slap AI on their portfolio and get access to AI-related investor money can be milked. On the consumer-facing front, they live off branding and if you're not using claude code, you might not associate the tool with Anthropic, which means losing publicity that drives API sales. So if they decide to raise prices in whatever shape or form then that's fine.Arbitrary restrictions do play a role for my own purchasing decisions though. Arbitrary restrictions do play a role for my own purchasing decisions though. Also why would you create a throwaway for this question? You should never question anyone's route to privacy :) If you have to ask, it's probably not rage bait. I'm just too lazy to come up with a username. Instead, many, many websites (especially in the music industry) have some sort of funky API that you can only get access to if you have enough online clout. Very few are transparent about what "enough clout" even means or how much it'd cost you, and there's like an entire industry of third-party API resellers that cost like 10x more than if you went straight to the source. Though, in this case, you get free API access to the model. I just think that OAI/Anthropic will try to keep both types of users locked into their walled garden via the UI.The APIs may have a future, but at our own peril and zero guarantees. The APIs may have a future, but at our own peril and zero guarantees. There is nothing here stopping cambridge analytica from doing this again, they will provide whatever details needed. But a small pre launch personal project work that might use a facebook publishing application can't be developed or tested without first going through all the bureaucracy.Nevermind the non profit 'free' application you might want to create on the FB platform, lets say a share chrome extension "Post to my FB", for personal use, you can't do this because you can't create an application without a company and IVA/TAX documents. It's hostile imo.Before, you could create an app, link your ToS, privacy policy etc, verify your domain via email, and then if users wanted to use your application they would agree, this is how a lot of companies still do it. I'm actually not sure why FB do this specifically. Nevermind the non profit 'free' application you might want to create on the FB platform, lets say a share chrome extension "Post to my FB", for personal use, you can't do this because you can't create an application without a company and IVA/TAX documents. It's hostile imo.Before, you could create an app, link your ToS, privacy policy etc, verify your domain via email, and then if users wanted to use your application they would agree, this is how a lot of companies still do it. I'm actually not sure why FB do this specifically. Before, you could create an app, link your ToS, privacy policy etc, verify your domain via email, and then if users wanted to use your application they would agree, this is how a lot of companies still do it. I'm actually not sure why FB do this specifically. Before you can sign up to build a WhatsApp bot, you need to jump through a million hoops, and after that, every automated message template must be vetted by Meta before it can be sent out, apple style.I'm glad of this, because unlike SMS and other messaging platforms, WhatsApp is spam free and a pleasure to use. I'm glad of this, because unlike SMS and other messaging platforms, WhatsApp is spam free and a pleasure to use. At least here in Italy whatsapp is a spam house unless you actively update the default privacy settings in-app. There is no discernable difference between SMS and WhatsApp to spammers. But the real issue is that these companies, once they have any market leverage, do things in their best interest to protect the little bit of moat they've acquired. Spotify in particular is just patently the very worst. It feels like their entire ecosystem has only ever gone downhill. https://web.archive.org/web/20141104154131/https://gigaom.co...But also, I feel like this broad repulsive trend is such an untenable position now that AI is here. Trying to make your app an isolated disconnected service is a suicide pact. And they are not going to be stopped even if you do try to control terms!Were I a smart engaged company, I'd be trying to build WebMCP access as soon as possible. Adoption will be slow, this isn't happening fast, but people who can mix human + agent activity on your site are going to be delighted by the experience, and that you will spread!WebMCP is better IMHO than conventional APIs because it layers into the experience you are already having. It's not a separate channel; it can build and use the session state of your browsing to do the things. Trying to make your app an isolated disconnected service is a suicide pact. And they are not going to be stopped even if you do try to control terms!Were I a smart engaged company, I'd be trying to build WebMCP access as soon as possible. Adoption will be slow, this isn't happening fast, but people who can mix human + agent activity on your site are going to be delighted by the experience, and that you will spread!WebMCP is better IMHO than conventional APIs because it layers into the experience you are already having. It's not a separate channel; it can build and use the session state of your browsing to do the things. Adoption will be slow, this isn't happening fast, but people who can mix human + agent activity on your site are going to be delighted by the experience, and that you will spread!WebMCP is better IMHO than conventional APIs because it layers into the experience you are already having. It's not a separate channel; it can build and use the session state of your browsing to do the things. WebMCP is better IMHO than conventional APIs because it layers into the experience you are already having. It's not a separate channel; it can build and use the session state of your browsing to do the things. I totally understand that I should not reuse my own account to provide services to others, as direct API usage is the obvious choice here, but this is a different case.I am currently developing something that would be the perfect fit for this OAuth based flow and I find it quite frustrating that in most cases I cannot find a clear answer to this question. I don't even know who I would be supposed to contact to get an answer or discuss this as an independent dev.EDIT: Some answers to my comment have pointed out that the ToS of Anthropic were clear, I'm not saying they aren't if taken in a vacuum, yet in practice even after this being published some confusion remained online, in particular regarding wether OAuth token usage was still ok with the Agent SDK for personal usage. Also, I am very interested about the stance of other companies on this subject.Maybe I am being overly cautious here but I want to be clear that this is just my personal opinion and me trying to understand what exactly is allowed or not. This is not some business or legal advice. I am currently developing something that would be the perfect fit for this OAuth based flow and I find it quite frustrating that in most cases I cannot find a clear answer to this question. I don't even know who I would be supposed to contact to get an answer or discuss this as an independent dev.EDIT: Some answers to my comment have pointed out that the ToS of Anthropic were clear, I'm not saying they aren't if taken in a vacuum, yet in practice even after this being published some confusion remained online, in particular regarding wether OAuth token usage was still ok with the Agent SDK for personal usage. Also, I am very interested about the stance of other companies on this subject.Maybe I am being overly cautious here but I want to be clear that this is just my personal opinion and me trying to understand what exactly is allowed or not. This is not some business or legal advice. EDIT: Some answers to my comment have pointed out that the ToS of Anthropic were clear, I'm not saying they aren't if taken in a vacuum, yet in practice even after this being published some confusion remained online, in particular regarding wether OAuth token usage was still ok with the Agent SDK for personal usage. Also, I am very interested about the stance of other companies on this subject.Maybe I am being overly cautious here but I want to be clear that this is just my personal opinion and me trying to understand what exactly is allowed or not. This is not some business or legal advice. Maybe I am being overly cautious here but I want to be clear that this is just my personal opinion and me trying to understand what exactly is allowed or not. This is not some business or legal advice. Subscriptions are for first-party products (claude.com, mobile and desktop apps, Claude Code, editor extensions, Cowork).Everything else must use API billing. But they're stating you can only use your subscription for your personal usage, not someone else's for their usage in your product.I honestly think they're being short sighted not just giving a "3rd party quota" since they already show users like 4 quotas.If the fear is 3rd party agents screwing up the math, just make it low enough for entry level usage. I suspect 3rd party token usage is bi-modal where some users just need enough to kick tires, but others are min-maxing for how mamy tokens they can burn as if that's its own reward I honestly think they're being short sighted not just giving a "3rd party quota" since they already show users like 4 quotas.If the fear is 3rd party agents screwing up the math, just make it low enough for entry level usage. I suspect 3rd party token usage is bi-modal where some users just need enough to kick tires, but others are min-maxing for how mamy tokens they can burn as if that's its own reward I suspect 3rd party token usage is bi-modal where some users just need enough to kick tires, but others are min-maxing for how mamy tokens they can burn as if that's its own reward Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service. I built a quick thing to download YouTube videos and transcribe them using with whisper, but it kind of feels clunky to summarize them using the claude CLI, even though that works. These kinds of business decisions show how these $200.00 subscriptions for their slot/infinite jest machines basically light that $200.00 on fire, and in general how unsustainable these business models are.Can't wait for it all to fail, they'll eventually try to get as many people to pay per token as possible, while somehow getting people to use their verbose antigentic tools that are able to inflate revenue through inefficient context/ouput shenanigans. I used Claude back when API per token pricing was the only option and it was bad for all the usual reasons pay-per-use sucks compared to flat billing: you're constantly thinking about cost. Like trying to watch a Netflix video with a ticker in the corner counting up the cents you owe them.I don't understand your claim that they want people paying per token - the subscription is the opposite of that, and it also has upsides for them as a business since most people don't saturate the usage limits, and the business gets to stuff a bunch of value-adds on a bundle offering which is generally a more lucrative and enticing consumer pricing model. I don't understand your claim that they want people paying per token - the subscription is the opposite of that, and it also has upsides for them as a business since most people don't saturate the usage limits, and the business gets to stuff a bunch of value-adds on a bundle offering which is generally a more lucrative and enticing consumer pricing model. Subscription- oh well, let's try again with a different promptPay per use- I just wasted money, this product sucksEven if it is less common than not, it has an outsized impact on how people feel using it. I expect some big falls from 10 figure businesses in the next year or two as they realize this is impossible. They've built an industry on the backs of gambling addicts and dopamine feins (I'm generalizing but this is a thing with LLM users (just read vibe coders posts on twitter, they're slot machine users). Ask sports betting operators from back in 2019-2022 how it worked out for them when they tried to give out 1-2k a year to attract new customers, and then realized their customers will switch platforms in an instant they see a new shiny offer. Look up the Fanduel Founders "exit" for an insight into this.They have to eventually stop catering to the slot machine users, which are generally paying for these hugely lossy flat rate subscriptions, and somehow get them used to a different type of payment model, or cater strictly to enterprise... Which also aren't going to tolerate paying 20k a month in tokens per developer, is my guess.... Lots of delicate pricing problems to figure out for all these companies. Which also aren't going to tolerate paying 20k a month in tokens per developer, is my guess.... Lots of delicate pricing problems to figure out for all these companies. But I already own my gaming PC that can run local models, and electricity is cheap. The issue with Claude Code is it's not at all obvious how any given task or query translates to cost. I was finding some days I spent very little and other days cost a fortune despite what seemed to me to be similar levels of usage. https://github.com/rivet-dev/sandbox-agent/tree/main/gigacod... [I saw this inShow HN: Gigacode – Use OpenCode's UI with Claude Code/Codex/Amp] (https://news.ycombinator.com/item?id=46912682)This can make Opencode work with Claude code and the added benefit of this is that Opencode has a Typescript SDK to automate and the back of this is still running claude code so technically should work even with the new TOS?So in the case of the OP. Maybe Opencode TS SDK <-> claude code (using this tool or any other like this) <-> It uses the oauth sign in option of Claude code users?Also, zed can use the ACP protocol itself as well to make claude code work iirc. So is using zed with CC still allowed?> I don't see how they can get more clear about this, considering they have repeatedly answered it the exact same way.This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. Maybe Opencode TS SDK <-> claude code (using this tool or any other like this) <-> It uses the oauth sign in option of Claude code users?Also, zed can use the ACP protocol itself as well to make claude code work iirc. So is using zed with CC still allowed?> I don't see how they can get more clear about this, considering they have repeatedly answered it the exact same way.This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. Maybe Opencode TS SDK <-> claude code (using this tool or any other like this) <-> It uses the oauth sign in option of Claude code users?Also, zed can use the ACP protocol itself as well to make claude code work iirc. So is using zed with CC still allowed?> I don't see how they can get more clear about this, considering they have repeatedly answered it the exact same way.This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. Also, zed can use the ACP protocol itself as well to make claude code work iirc. So is using zed with CC still allowed?> I don't see how they can get more clear about this, considering they have repeatedly answered it the exact same way.This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. > I don't see how they can get more clear about this, considering they have repeatedly answered it the exact same way.This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. This is confusing quite frankly, there's also the claude agent sdk thing which firloop and others talked about too. You can't use Claude OAuth tokens for anything. Any solution that exists worked because it pretended/spoofed to be Claude Code. Same for Gemini (Gemini CLI, Antigravity)Codex is the only one that got official blessing to be used in OpenClaw and OpenCode, and even that was against the ToS before they changed their stance on it. Codex is the only one that got official blessing to be used in OpenClaw and OpenCode, and even that was against the ToS before they changed their stance on it. Use it when you want a deep integration inside your own product. But I believe OpenAI does let you use their subscription in third parties, so not an issue anyway. A third-party tool may be less efficient in saving costs (I have heard many of them don't hit Anthropic LLMs' caches as well).Would you be willing to pay more for your plan, to subsidize the use of third-party tools by others?---Note, afaik, Anthropic hasn't come out and said this is the reason, but it fits.Or, it could also just be that the LLM companies view their agent tools as the real moat, since the models themselves aren't. If OpenCode or whatever burns through your limits faster, are you likely to place the blame on OpenCode?Maybe a good analogy was when DoorDash/GrubHub/Uber Eats/etc signed up restaurants to their system without their permission. When things didn't go well, the customers complained about the restaurants, even though it wasn't their fault, because they chose not to support delivery at scale.Second, flat-rate pricing, unlike API pricing, is the same for cached vs uncached iirc, so even if total token limits are the same, less caching means higher costs. If OpenCode or whatever burns through your limits faster, are you likely to place the blame on OpenCode?Maybe a good analogy was when DoorDash/GrubHub/Uber Eats/etc signed up restaurants to their system without their permission. When things didn't go well, the customers complained about the restaurants, even though it wasn't their fault, because they chose not to support delivery at scale.Second, flat-rate pricing, unlike API pricing, is the same for cached vs uncached iirc, so even if total token limits are the same, less caching means higher costs. When things didn't go well, the customers complained about the restaurants, even though it wasn't their fault, because they chose not to support delivery at scale.Second, flat-rate pricing, unlike API pricing, is the same for cached vs uncached iirc, so even if total token limits are the same, less caching means higher costs. Second, flat-rate pricing, unlike API pricing, is the same for cached vs uncached iirc, so even if total token limits are the same, less caching means higher costs. Probably, but I get your point that your average user would blame Anthropic instead.> even if total token limits are the same, less caching means higher costsNot really, flat-rate pricing simply gives you a fixed token allotment, so less caching means you consume your 5-hour/weekly allotment faster. > even if total token limits are the same, less caching means higher costsNot really, flat-rate pricing simply gives you a fixed token allotment, so less caching means you consume your 5-hour/weekly allotment faster. Not really, flat-rate pricing simply gives you a fixed token allotment, so less caching means you consume your 5-hour/weekly allotment faster. They control how fast your subscription tokens are burned They control how fast your subscription tokens are burned They control how fast your subscription tokens are burned Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service. I have never noticed there are people who interpret it that way. I'm sure you can use context clues to figure this one out. Advertised usage limits for Pro and Max plans assume ordinary, individual usage of Claude Code and the Agent SDK ""That tool clearly falls under ordinary individual use of Claude code. Perfectly ordinary individual usage.https://yepanywhere.com/sdk-auth-clarification.htmlThe TOS are confusing because just below that section it talks about authentication/credential use. If an app starts reading api keys / credentials, that starts falling into territory where they want a hard line no. Advertised usage limits for Pro and Max plans assume ordinary, individual usage of Claude Code and the Agent SDK ""That tool clearly falls under ordinary individual use of Claude code. Perfectly ordinary individual usage.https://yepanywhere.com/sdk-auth-clarification.htmlThe TOS are confusing because just below that section it talks about authentication/credential use. If an app starts reading api keys / credentials, that starts falling into territory where they want a hard line no. That tool clearly falls under ordinary individual use of Claude code. Perfectly ordinary individual usage.https://yepanywhere.com/sdk-auth-clarification.htmlThe TOS are confusing because just below that section it talks about authentication/credential use. If an app starts reading api keys / credentials, that starts falling into territory where they want a hard line no. https://yepanywhere.com/sdk-auth-clarification.htmlThe TOS are confusing because just below that section it talks about authentication/credential use. If an app starts reading api keys / credentials, that starts falling into territory where they want a hard line no. The TOS are confusing because just below that section it talks about authentication/credential use. If an app starts reading api keys / credentials, that starts falling into territory where they want a hard line no. I can't find anything official from OpenAI, but they have worked with the OpenCode people to support using your ChatGPT subscription in OpenCode. Opus has gone down the hill continously in the last week (and before you start flooding with replies, I've been testing opus/codex in parallel for the last week, I've plenty of examples of Claude going off track, then apologising, then saying "now it's all fixed!" and then only fixing part of it, when codex nailed at the first shot).I can accept specific model limits, not an up/down in terms of reliability. And don't even let me get started on how bad Claude client has become. Others are finally catching up and gpt-5.3-codex is definitely better than opus-4.6Everyone else (Codex CLI, Copilot CLI etc...) is going opensource, they are going closed. I can accept specific model limits, not an up/down in terms of reliability. And don't even let me get started on how bad Claude client has become. Others are finally catching up and gpt-5.3-codex is definitely better than opus-4.6Everyone else (Codex CLI, Copilot CLI etc...) is going opensource, they are going closed. This hostile behaviour is just the last drop. It seems like they currently have a lot of false positives: https://github.com/openai/codex/issues?q=High%20risk That pattern is people complaining that a particular model has degraded in quality of its responses over time or that it has been “nerfed” etc.Although the models may evolve, and the tools calling them may change, I suspect a huge amount of this is simply confirmation bias. Although the models may evolve, and the tools calling them may change, I suspect a huge amount of this is simply confirmation bias. My brain trailed off after "won't be long enough to even finish"... ps: imafish may only be a fan of But if people really like Codex better, maybe I'll try it. I've been trying not to pay for 2 subscriptions at once but it might be worth a test. It does seem to spend a lot more time “thinking” before giving what feels like equivalent results, most of the time.Probably eats into the gambling-style adrenaline cycles. Opus 4.6 wrote me a working macos application.Codex wrote me a html + css mockup of a macos application that didn't even look like a macos application at all.Opus 4.5 was fine, but I feel that 4.6 is more often on the money on its implementations than 4.5 was. Codex wrote me a html + css mockup of a macos application that didn't even look like a macos application at all.Opus 4.5 was fine, but I feel that 4.6 is more often on the money on its implementations than 4.5 was. I generally don't like the way codex approaches coding itself so I just feed its review comments back in to Claude Code and off we go. In my experience, two different models together works much better than one, that's why this subscription banning is distressing. And there's a good reason the most "famous" vibe coders, including the OpenClaw creator all moved to Codex, it's just better.Claude writes a lot more code to do anything, tons of redundent code, repeated code etc. Codex is only model I've seen which occasionally removes more code than it writes. Codex is only model I've seen which occasionally removes more code than it writes. The latest versions of claude code have been freezing and then crashing while waiting on long running commands. I have a feeling Anthropic might be in for an extremely rude awakening when that happens, and I don't think it's a matter of “if” anymore. This was inevitable, I do not understand why we trust these companies, ever. It will be real interesting if the haters are right and this technology is not the breakthrough the investors assume it to be AFTER it is already sewn into everyone's work flows. Everyone keeps talking about how jobs will be displaced, yet few are asking what happens when a dependency is swept out from underneath the industry as a whole if/when this massive gamble doesn't pay off.Whatever. I am squawking into the void as we just repeat history. I am squawking into the void as we just repeat history. First, we are not talking about a cheap service here. We are talking about a monthly subscription which costs 100 USD or 200 USD per month, depending on which plan you choose.Second, it's like selling me a pizza and pretending I only eat it while sitting at your table. I want to eat the pizza at home. I want to eat the pizza at home. Claude has gotten a lot of popular media attention in the last few weeks, and the influx of users is constraining compute/memory on an already compute heavy model. So you get all the suspected "tricks" like quantization, shorter thinking, KV cache optimizations.It feels like the same thing that happened to Gemini 3, and what you can even feel throughout the day (the models seem smartest at 12am).Dario in his interview with dwarkesh last week also lamented the same refrain that other lab leaders have: compute is constrained and there are big tradeoffs in how you allocate it. It feels like the same thing that happened to Gemini 3, and what you can even feel throughout the day (the models seem smartest at 12am).Dario in his interview with dwarkesh last week also lamented the same refrain that other lab leaders have: compute is constrained and there are big tradeoffs in how you allocate it. Dario in his interview with dwarkesh last week also lamented the same refrain that other lab leaders have: compute is constrained and there are big tradeoffs in how you allocate it. At least weekly I run a set of prompts to compare codex/claude against each other. This is quite easy the prompt sessions are just text files that are saved.The problem is doing it enough for statistical significance and judging the output as better or not. I'm working through a refactor and I explicitly told it to use a block (as in Ruby Blocks) and it completely overlooked that. What is interesting is that OpenAI and GitHub seem to be taking the opposite approach with Copilot/OpenCode, essentially treating third-party tool access as a feature that increases subscription stickiness. Different bets on whether the LTV of a retained subscriber outweighs the marginal inference cost.Would not be surprised if this converges eventually. Either Anthropic opens up once their margins improve, or OpenAI tightens once they realize the arbitrage is too expensive at scale. Would not be surprised if this converges eventually. Either Anthropic opens up once their margins improve, or OpenAI tightens once they realize the arbitrage is too expensive at scale. I'm rather certain, though cannot prove it, that buying the same tokens would cost at least 10x more if bought from API. Anecdotally, my cursor team usage was getting to around 700$ / month. After switching to claude code max, I have so far only once hit the 3h limit window on the 100$ sub.What Im thinking is that Anthropic is making loss with users who use it a lot, but there are a lot of users who pay for max, but don't actually use it.With the recent improvements and increase of popularity in projects like OpenClaw, the number of users that are generating loss has probably massively increased. What Im thinking is that Anthropic is making loss with users who use it a lot, but there are a lot of users who pay for max, but don't actually use it.With the recent improvements and increase of popularity in projects like OpenClaw, the number of users that are generating loss has probably massively increased. With the recent improvements and increase of popularity in projects like OpenClaw, the number of users that are generating loss has probably massively increased. edit: My $40/month subscription used $662 worth of API credits. I don't entirely mind, and am just considering it an even better work:life balance, but if this is $200 worth of queries, then all I can say is LOL. I'm messing around on document production, I can't imagine being on a crunch facing a deadline or dealing with a production issue and 1) seeing some random fuck-up eat my budget with no take backs (‘sure thing, I'll make a custom docx editor to open that…'), 2) having to explain to my boss why Thursday cost $500 more than expected because of some library mismatch, or 3) trying to decide whether we're gonna spend or wait while stressing some major issue (the LLM got us in it, so we kinda need the LLM to get us out).That's a lot of extra shizz on top of already tricky situations. API limits are infinite but you'd blow through $20 of usage in a maybe 1 hours or less of intense Opus use.The subscription at $20/mo (or $200) allows for vastly more queries than $20 would buy you via API but you are constrained by hourly/weekly limits.The $20/mo sub user will take a lot longer to complete a high token count task (due to start/stop) BUT they will cap their costs. The subscription at $20/mo (or $200) allows for vastly more queries than $20 would buy you via API but you are constrained by hourly/weekly limits.The $20/mo sub user will take a lot longer to complete a high token count task (due to start/stop) BUT they will cap their costs. The $20/mo sub user will take a lot longer to complete a high token count task (due to start/stop) BUT they will cap their costs. Their bet is that most people will not fill up 100% of their weekly usage for 4 consecutive weeks of their monthly plan, because they are humans and the limits impede long running tasks during working hours. I dont like it either, but its not an unreasonable restriction. obviously this WILL happen eventually when they need to pay for things. Banning third-party tools has nothing to do with rate limits. They're trying to position themselves as the Apple of AI companies -a walled garden. They may soon discover that screwing developers is not a good strategy.They are not 10× better than Codex; on the contrary, in my opinion Codex produces much better code. Forcing people to use ONLY a broken Claude Code UX with a subscription only ensures they loose advantage they had. Forcing people to use ONLY a broken Claude Code UX with a subscription only ensures they loose advantage they had. Google AI Pro is like $15/month for practically unlimited Pro requests, each of which take million tokens of context (and then also perform thinking, free Google search for grounding, inline image generation if needed). This includes Gemini CLI, Gemini Code Assist (VS Code), the main chatbot, and a bunch of other vibe-coding projects which have their own rate limits or no rate limits at all.It's crazy to think this is sustainable. Google has made custom AI chips for 11 years — since 2015 — and inference costs them 2-5x less than it does for every other competitor.The landmark paper that invented the techniques behind ChatGPT, Claude and modern AI was also published by Google scientists 9 years ago.That's probably how they can afford it. The landmark paper that invented the techniques behind ChatGPT, Claude and modern AI was also published by Google scientists 9 years ago.That's probably how they can afford it. Google already has a huge competitive advantage because they have more data than anyone else, bundle Gemini in each android to siphon even more data, and the android platform. The TPUs truly make me believe there actually could be a sort of monopoly on LLMs in the end, even though there are so many good models with open weights, so little (technical) reasons to create software that only integrates with Gemini, etc.Google will have a lion‘s share of inferring I believe. OpenAI and Claude will have a very hard time fighting this. Google will have a lion‘s share of inferring I believe. OpenAI and Claude will have a very hard time fighting this. 5h allowance is somewhere between 50M-100M tokens from what I can tell.On 200$ claude code plan you should be burning hundreds of millions of token per day to make anthropic hurt.IMHO subscription plans are totally banking on many users underusing them. Also LLM providers dont like to say exact numbers (how much you get , etc) On 200$ claude code plan you should be burning hundreds of millions of token per day to make anthropic hurt.IMHO subscription plans are totally banking on many users underusing them. Also LLM providers dont like to say exact numbers (how much you get , etc) IMHO subscription plans are totally banking on many users underusing them. Also LLM providers dont like to say exact numbers (how much you get , etc) Most people lose money on their gym subscription, but the convenience takes us. The interesting question is: In what scenario do you see any of the players as being able to stop spending ungodly amounts for R&D and hardware without losing out to the competitors? Why do people keep saying inference is cheap if they're losing so much money from it? You can buy a GPU that's been used to mine bitcoin for 5 years with zero downtime, and as long as it's been properly taken care of (or better, undervolted), that GPU functions the exact same as a 5 year old GPU in your PC. Probably even better.GPUs are rated to do 100%, all the time. GPUs are rated to do 100%, all the time. The only reason they're "perishable" is because of the GPU arms race, where renewing them every 5 years is likely to be worth the investment for the gains you make in power efficiency.Do you think Google has a pile of millions of older TPUs they threw out because they all failed, when chips are basically impossible to recycle ? Do you think Google has a pile of millions of older TPUs they threw out because they all failed, when chips are basically impossible to recycle ? "Synthesizing drugs is cheap - just a few dollars per million pills. "There's plenty of legit criticisms of this business model and Anthropic, but pointing out that R&D companies sink money into research and then charge more than the marginal cost for the final product, isn't one of them. "Synthesizing drugs is cheap - just a few dollars per million pills. "There's plenty of legit criticisms of this business model and Anthropic, but pointing out that R&D companies sink money into research and then charge more than the marginal cost for the final product, isn't one of them. There's plenty of legit criticisms of this business model and Anthropic, but pointing out that R&D companies sink money into research and then charge more than the marginal cost for the final product, isn't one of them. My point was simpler: they're almost certainly not losing money on subscriptions because of inference. They're competing with companies like Kimi and DeepSeek that also spend heavily on R&D but release strong models openly. That means anyone can run inference and customers can use it without paying for bundled research costs.Training frontier models takes months, costs billions, and the model is outdated in six months. I just don't see how a closed, subscription-only model reliably covers that in the long run, especially if you're tightening ecosystem access at the same time. They're competing with companies like Kimi and DeepSeek that also spend heavily on R&D but release strong models openly. That means anyone can run inference and customers can use it without paying for bundled research costs.Training frontier models takes months, costs billions, and the model is outdated in six months. I just don't see how a closed, subscription-only model reliably covers that in the long run, especially if you're tightening ecosystem access at the same time. I just don't see how a closed, subscription-only model reliably covers that in the long run, especially if you're tightening ecosystem access at the same time. They can totally lose money on subscriptions despite the costs of inference, because research costs have to be counted too. Of course they are losing money when you factor in R&D. If AI was truly this productive they wouldn't be struggling so hard to sell their wares. Enterprise products with sufficient market share and "stickiness", will not.For historical precedent, see the commercial practices of Oracle, Microsoft, Vmware, Salesforce, at the height of their power. The software is free (citation: Cuda, nvcc, llvm, olama/llama cpp, linux, etc)The hardware is *not* getting cheaper (unless we're talking a 5+ year time) as most manufacturers are signaling the current shortages will continue ~24 months. The hardware is *not* getting cheaper (unless we're talking a 5+ year time) as most manufacturers are signaling the current shortages will continue ~24 months. Yes, that's the time I'm talking about.You also had a blip with increasing hard disk prices when Thailand flooded a few years ago. We see vendors reducing memory in new smart phones in 2026 vs 2025 for example.At least for the moment falling consumer tech hardware prices are over. At least for the moment falling consumer tech hardware prices are over. I recently encountered this randomly -- knives are apparently one of the few products that nearly every household has needed since antiquity, and they have changed fairly little since the bronze age, so they are used by economists as a benchmark that can span centuries.Source: it was an aside in a random economics conversation with charGPT (grain of salt? ).There is no practical upshot here, but I thought it was cool. Source: it was an aside in a random economics conversation with charGPT (grain of salt? ).There is no practical upshot here, but I thought it was cool. There is no practical upshot here, but I thought it was cool. Not to mention the advances in composite materials for handles.Then you need to look at substitute goods and the what people actually used knives for.A huge amount of the demand for knives evaporated thanks to societal changes and substitute goods like forks. Knives like that exist today but they're not something every household has or needs.This is a good example of why learning from ChatGPT is dangerous. This is a story that sounds very plausible at first glance, but doesn't make sense once you dig in. Not to mention the advances in composite materials for handles.Then you need to look at substitute goods and the what people actually used knives for.A huge amount of the demand for knives evaporated thanks to societal changes and substitute goods like forks. Knives like that exist today but they're not something every household has or needs.This is a good example of why learning from ChatGPT is dangerous. This is a story that sounds very plausible at first glance, but doesn't make sense once you dig in. Then you need to look at substitute goods and the what people actually used knives for.A huge amount of the demand for knives evaporated thanks to societal changes and substitute goods like forks. Knives like that exist today but they're not something every household has or needs.This is a good example of why learning from ChatGPT is dangerous. This is a story that sounds very plausible at first glance, but doesn't make sense once you dig in. A huge amount of the demand for knives evaporated thanks to societal changes and substitute goods like forks. Knives like that exist today but they're not something every household has or needs.This is a good example of why learning from ChatGPT is dangerous. This is a story that sounds very plausible at first glance, but doesn't make sense once you dig in. This is a story that sounds very plausible at first glance, but doesn't make sense once you dig in. It was sticker price of $33,000 adjusted for inflation:https://en.wikipedia.org/wiki/Ford_Taurus_%28second_generati...I don't think it would even feel safe to drive at all compared to what we have got use to with modern cars. No cell phone of course to call anyone.These were the mythic "good ol days". No cell phone of course to call anyone.These were the mythic "good ol days". No cell phone of course to call anyone.These were the mythic "good ol days". I also think we're, as ICs, being given Bentleys meanwhile they're trying to invent Waymos to put us all out of work.Humans are the cost center in their world model. Humans are the cost center in their world model. Finance 101 tldr explanation: The contribution margin (= price per token -variable cost per token ) this is positiveProfit (= contribution margin x cuantity- fix cost) i will not be as bullish to say they will no colapse (0 idear how much real debt and commitments they have, if after the bubble pop spending fall shraply, or a new deepseek moment) but this sound like good trajectory (all things considered) i heavily doubt the 380 billions in valuation"this is how much is spendeed in developers between $659 billion and $737 billion. If the answer is not yes, then they are making money on inference. The sounds like a confession that claude code is somewhat wasteful at token use. What a PR nightmare, on top of an already bad week. I've seen 20+ people on X complaining about this and the related confusion. It's slow to use and so far behind the Codex app that it's embarassing.- Claude Code is buggy has hell and I think I've never used a CLI tool that consume so much memory and CPU. Let's not talk about the feature parity with other agents.- Claude Agent SDK is poorly documented, half finished, and is just thin wrapper around a CLI tool…Oh and none of this is open source, so I can do nothing about it.My only option to stay with their model is to build my own tool. And now I discover that using my subscription with the Agent SDK is against the term of use?I'm not going to pay 500 USD of API credits every months, no way. - Claude Code is buggy has hell and I think I've never used a CLI tool that consume so much memory and CPU. Let's not talk about the feature parity with other agents.- Claude Agent SDK is poorly documented, half finished, and is just thin wrapper around a CLI tool…Oh and none of this is open source, so I can do nothing about it.My only option to stay with their model is to build my own tool. And now I discover that using my subscription with the Agent SDK is against the term of use?I'm not going to pay 500 USD of API credits every months, no way. And now I discover that using my subscription with the Agent SDK is against the term of use?I'm not going to pay 500 USD of API credits every months, no way. Oh and none of this is open source, so I can do nothing about it.My only option to stay with their model is to build my own tool. And now I discover that using my subscription with the Agent SDK is against the term of use?I'm not going to pay 500 USD of API credits every months, no way. And now I discover that using my subscription with the Agent SDK is against the term of use?I'm not going to pay 500 USD of API credits every months, no way. I'm not going to pay 500 USD of API credits every months, no way. It seems to me that other CLI agents are quite far from Claude Code in this regard. What do you mean feature parity with other agents? It seems to me that other CLI agents are quite far from Claude Code in this regard. FWIW this aligns completely with the LLM ethos. We are heading toward a $1000/month model just to use LLMs in the cloud. This policy is souring the relationship, and basically saying that Claude isn't a keeper.I'll keep my eye-watering sub for now because it's still working out, but this ensures I won't feel bad about leaving when the time comes.Update: yes yes, API, I know. We're in the part of the market cycle where everyone fights for marketshare by selling dollar bills for 50 cents.When a winner emerges they'll pull the rug out from under you and try to wall off their garden.Anthropic just forgot that we're still in the "functioning market competition" phase of AI and not yet in the "unstoppable monopoly" phase. When a winner emerges they'll pull the rug out from under you and try to wall off their garden.Anthropic just forgot that we're still in the "functioning market competition" phase of AI and not yet in the "unstoppable monopoly" phase. Anthropic just forgot that we're still in the "functioning market competition" phase of AI and not yet in the "unstoppable monopoly" phase. all closed AI model providers will stop selling APIs in the next 2-3 years. Only open models will be available via APIs (…) Closed model providers are trying to build non-commodity capabilities and they need great UIs to deliver those. "~ https://vintagedata.org/blog/posts/model-is-the-product A. Doria> new Amp Free (10$) access is also closed up since of last night ~ https://vintagedata.org/blog/posts/model-is-the-product A. Doria> new Amp Free (10$) access is also closed up since of last night > new Amp Free (10$) access is also closed up since of last night I only use LLMs through OpenRouter and switch somewhat randomly between frontier models; they each have some amount of personality but I wouldn't mind much if half of them disappeared overnight, as long as the other half remained available. I imagine the split will look a lot like b2b vs b2c in other technologies, b2b customers tend to be willing to pay for tech when it offers a competitive advantage, reduces their operating costs etc. (Not quite "every", but outside of tech, most professional workplaces don't support ad blocking or Kagi.) I just don't think this is the same as Google.Of course I could be entirely mistaken and there could emerge a single winner I just don't think this is the same as Google.Of course I could be entirely mistaken and there could emerge a single winner Every time a model gets used by esoteric use cases, it gets more training data (that a decentralized open weight model doesn't get) and it starts developing its moat. They bundled it with PC hw and the vast majority of apps only ever got published for windows, and this over decades (one would argue it's still true).The starting point for LLMs is very different. Who would publish today a software that only integrates with chatGPT? The starting point for LLMs is very different. Who would publish today a software that only integrates with chatGPT? More directly, enterprise-software stocks are getting hammered because AI offers them competition. (For those unaware, AWS doesn't have a VM monopoly, and the market dynamics seem similar) we don't, we have about 3 operating systems that have the decades of hardware and software compatibility that makes them widely usable. They're the most complex and complicated things we've built. This is the least defensible piece of software there ever has been. And if the frontier continues favouring centralised solutions, they'll get it. Just looking at how much Claude complains about me not paying for SSO-tier subscriptions to data tools when they work perfectly fine in a browser is starting to make running a slower, less-capable model locally competitive with it in some research contexts. It almost makes me feel sorry for Dario despite fundamentally disliking him as a person. First of all, custom harness parallel agent people are so far from the norm, and certainly not on the $20 plan, which doesn't even make sense because you'd hit token limit in about 90 seconds.Second, token limits. If I'm paying a blistering monthly fee, I should be able to use up to the limit.Now I know you've got a clear view of the typical user, but FWIW, I'm just an aging hacker using CC to build some personal projects (feeling modern ofc) but still driving, no yolo or gas town style. I've reached the point where I have a nice workflow, and CC is pretty decent, but it feels like it's putting on weight and adding things I don't want or need.I think LLMs are an exciting new interface to computers, but I don't want to be tied to someone else's idea of a client, especially not one that's changing so rapidly. I'd like to roll my own client to interface with the model, or maybe try out some other alternatives, but that's against the TOS, because: reasons.And no, I'm not interested in paying metered corporate rates for API access. I pay for a Max account, it's expensive, but predictable.The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. If I'm paying a blistering monthly fee, I should be able to use up to the limit.Now I know you've got a clear view of the typical user, but FWIW, I'm just an aging hacker using CC to build some personal projects (feeling modern ofc) but still driving, no yolo or gas town style. I've reached the point where I have a nice workflow, and CC is pretty decent, but it feels like it's putting on weight and adding things I don't want or need.I think LLMs are an exciting new interface to computers, but I don't want to be tied to someone else's idea of a client, especially not one that's changing so rapidly. I'd like to roll my own client to interface with the model, or maybe try out some other alternatives, but that's against the TOS, because: reasons.And no, I'm not interested in paying metered corporate rates for API access. I pay for a Max account, it's expensive, but predictable.The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. Now I know you've got a clear view of the typical user, but FWIW, I'm just an aging hacker using CC to build some personal projects (feeling modern ofc) but still driving, no yolo or gas town style. I've reached the point where I have a nice workflow, and CC is pretty decent, but it feels like it's putting on weight and adding things I don't want or need.I think LLMs are an exciting new interface to computers, but I don't want to be tied to someone else's idea of a client, especially not one that's changing so rapidly. I'd like to roll my own client to interface with the model, or maybe try out some other alternatives, but that's against the TOS, because: reasons.And no, I'm not interested in paying metered corporate rates for API access. I pay for a Max account, it's expensive, but predictable.The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. I'd like to roll my own client to interface with the model, or maybe try out some other alternatives, but that's against the TOS, because: reasons.And no, I'm not interested in paying metered corporate rates for API access. I pay for a Max account, it's expensive, but predictable.The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. And no, I'm not interested in paying metered corporate rates for API access. I pay for a Max account, it's expensive, but predictable.The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. The issue is Anthropic is trying for force users into using their tool, but that's not going to work for something so generic as interfacing with an LLM. Some folks want emacs while others want vim, and there will never be a consensus on the best editor (it's nvim btw), because developers are opinionated and have strong preferences for how they interface with computers. I don't give a shit about Anthropic's credit liability, I just want the freedom to hack on my own client. That's not a principled stance about interface freedom, it's just wanting something for less than it costs.The nvim analogy doesn't land either. Maybe the API would be cheaper for me, I don't know. I'm just a normal user, not scheduling an agent army to blast at maximum 24/7.You don't need to explain what Anthropic is selling, I get it, but you're off-base claiming that I'm pretending about editor philosophy as cope. I think Anthropic is where they are today because they have a good model for coding, made popular by software folks who value things like editor choice and customization. Anthropic is free to do as they wish of course, as am I, but I'm displeased with their decision here, and voicing my opinion about it.If usage is truly constrained by the client not the server, I guess I can't argue that, but it still feels bad as an end user. But that seems harder to find these days, and most businesses seem intent on maximum extraction by any means possible. I might be wrong, but this feels like business move to build a consumer moat by controlling the interface, because consumers don't want the API. It's not in my best interests, which alienates me as a customer. You don't need to explain what Anthropic is selling, I get it, but you're off-base claiming that I'm pretending about editor philosophy as cope. I think Anthropic is where they are today because they have a good model for coding, made popular by software folks who value things like editor choice and customization. Anthropic is free to do as they wish of course, as am I, but I'm displeased with their decision here, and voicing my opinion about it.If usage is truly constrained by the client not the server, I guess I can't argue that, but it still feels bad as an end user. But that seems harder to find these days, and most businesses seem intent on maximum extraction by any means possible. I might be wrong, but this feels like business move to build a consumer moat by controlling the interface, because consumers don't want the API. It's not in my best interests, which alienates me as a customer. If usage is truly constrained by the client not the server, I guess I can't argue that, but it still feels bad as an end user. But that seems harder to find these days, and most businesses seem intent on maximum extraction by any means possible. I might be wrong, but this feels like business move to build a consumer moat by controlling the interface, because consumers don't want the API. It's not in my best interests, which alienates me as a customer. The only thing I've seen from him that I don't like is the "SWEs will be replaced" line (which is probably true and it's more that I don't like the factuality of it). Running locally is going to require a lot of memory, compute, and energy for the foreseeable future which makes it really hard to compete with ~$20/mo subscriptions. This could be used to adhere to Claude's TOS while still allowing the user to switch AI companies at a moment's notice.Right now there's limited customizability in this approach, but I think it's not far-fetched to see FAR more integrated solutions in the future if the lock-in trend continues. For example: one MCP that you can configure into a coding agent like Claude Code that overrides its entire behavior (tools, skills, etc.) Think something similar to the existing IntelliJ IDEA's MCP that gives a separate file edit tool, etc. than the one the agent comes with.Illustration of what i'm talking about:- You install Claude Code with no configuration- Then you install the meta-agent framework- With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) Right now there's limited customizability in this approach, but I think it's not far-fetched to see FAR more integrated solutions in the future if the lock-in trend continues. For example: one MCP that you can configure into a coding agent like Claude Code that overrides its entire behavior (tools, skills, etc.) Think something similar to the existing IntelliJ IDEA's MCP that gives a separate file edit tool, etc. than the one the agent comes with.Illustration of what i'm talking about:- You install Claude Code with no configuration- Then you install the meta-agent framework- With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) Illustration of what i'm talking about:- You install Claude Code with no configuration- Then you install the meta-agent framework- With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - You install Claude Code with no configuration- Then you install the meta-agent framework- With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - Then you install the meta-agent framework- With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - With one command the meta-agent MCP is installed in Claude Code, built-in tools are disabled via permissions override- You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - You access the meta-agent through a different UI (similar to vibe-kanban's web UI)- Everything you do gets routed directly to Claude Code, using your Claude subscription legally. (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) (Input-level features like commands get resolved by meta-agent UI before being sent to claude code)- Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - Claude Code must use the tools and skills directly from meta-agent MCP as instructed in the prompt, and because its own tools are permission denied (result: very good UI integration with the meta-agent UI)- This would also work with any other CLI coding agent (Codex, Gemini CLI, Copilot CLI etc.) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) should they start getting ideas of locking users in- If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) - If Claude Code rug-pulls subscription quotas, just switch to a competitor instantlyAll it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) All it requires is a CLI coding agent with MCP support, and the TOS allowing automatic use of its UI (disallowing that would be massive hypocrisy as the AI companies themselves make computer use agents that allow automatic use of other apps' UI) The second appears to be hitching my wagon to Mistral even though it's apparently nowhere as powerful or featureful as the big guys. But do you know how many times they've screwed me over? Not once.Maybe it's my use cases that make this possible. Maybe it's my use cases that make this possible. From a backend perspective, the subscription model creates perverse incentives. Third-party tools amplify this asymmetry.Anthropic's move is economically rational but strategically risky. Models are increasingly fungible - Gemini 3.1 and Claude 4.5 produce similar results for most tasks. Given how quickly open-source harnesses like pi have caught up, that's a bold bet. Anthropic's move is economically rational but strategically risky. Models are increasingly fungible - Gemini 3.1 and Claude 4.5 produce similar results for most tasks. Given how quickly open-source harnesses like pi have caught up, that's a bold bet. By forcing users onto Claude Code exclusively, they're betting their tooling moat is stronger than competitor models. Given how quickly open-source harnesses like pi have caught up, that's a bold bet. It's merely the hardware that should be charged for - which ought to drop in price if/when the demand for it rises. However, this is a bottleneck at the moment, and hard to see how it gets resolved amidst the current US environment on sanctioning anyone who would try. And i would also argue that the researchers doing this are built on shoulders of other public knowledge - things funded by public institutions with taxpayer money. And as a bonus, you can choose your harness. You don't have to suffer CC.And if something better appears tomorrow, you switch your model, while still using your harness of choice. But the big guys don't seem interested in this, maybe some lesser known model will carve out this space I shudder to think what the industry will look like if software development and delivery becomes like Youtubing, where the whole stack and monetization is funneled through a single company (or a couple) get to decide who gets how much money. As an independent dev I also unfortunately don't have investors backing me to subsidize inference for my subscription plan. very comparable to sonnet/opus although kimi isn't the best in coding. I think its a really great solid model overall and might just be worth it in your use case?Is the use case extremely coding intensive related (where even some minor improvement can matter for 10-100x cost) or just in general. Because if not, then I can recommend Kimi. Is the use case extremely coding intensive related (where even some minor improvement can matter for 10-100x cost) or just in general. Because if not, then I can recommend Kimi. Please correct me if you feel I'm wrong after reading it. Maybe they are not worth building at all then. And historically, embedded/OEM use cases always have different pricing models for a variety of reasons why.How is this any different than this long established practice? > Authentication and credential use> Claude Code authenticates with Anthropic's servers using OAuth tokens or API keys. These authentication methods serve different purposes:> OAuth authentication (used with Free, Pro, and Max plans) is intended exclusively for Claude Code and Claude.ai. Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service.> Developers building products or services that interact with Claude's capabilities, including those using the Agent SDK, should use API key authentication through Claude Console or a supported cloud provider. > Claude Code authenticates with Anthropic's servers using OAuth tokens or API keys. These authentication methods serve different purposes:> OAuth authentication (used with Free, Pro, and Max plans) is intended exclusively for Claude Code and Claude.ai. Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service.> Developers building products or services that interact with Claude's capabilities, including those using the Agent SDK, should use API key authentication through Claude Console or a supported cloud provider. Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service.> Developers building products or services that interact with Claude's capabilities, including those using the Agent SDK, should use API key authentication through Claude Console or a supported cloud provider. > Developers building products or services that interact with Claude's capabilities, including those using the Agent SDK, should use API key authentication through Claude Console or a supported cloud provider. https://x.com/i/status/2024212378402095389---On a different note, it's surprising that a company that size has to clarify something as important as ToS via X ---On a different note, it's surprising that a company that size has to clarify something as important as ToS via X On a different note, it's surprising that a company that size has to clarify something as important as ToS via X Seriously it feels like half of the EU parliament live on twitter. Plus it's not a real clarification in anyway. Even if it's posted on Mastodon or Github or anywhere, I highly doubt you can use it to defend yourself if you get banned from violating their ToS. I think what they want to achieve here is less "kill openclaw" or similar and more "keep our losses under control in general". And now they have a clear criteria to refer when they take action and a good bisection on whom to act on.In case your usage is high they would block / take action. Because if you have your max subscription and not really losing them money, why should they push you (the monopoly incentive sounds wrong with the current market). In case your usage is high they would block / take action. Because if you have your max subscription and not really losing them money, why should they push you (the monopoly incentive sounds wrong with the current market). But Opus is particularly good for "agent with a personality" applications, so it's what thousands of OpenClaw users go with, mostly via the OAuth token, because it's much cheaper than the API. They are literally alienating a large percentage of OpenClaw, NanoClaw, PicoClaw, customers because those customers will surely not be willing to pay API pricing, which is at least 6-10x Max Plan pricing (for my usage).This isn't too surprising to me since they probably have a direct competitor to openclaw et al in the works right now, but until then I am cancelling my subscription and porting my nanoclaw fork with mem0 integration to work with OpenAI instead.Thats not a “That'll teach ‘em” statement, it is just my own cost optimization. I am quite fond of Anthropic's coding models and might still subscribe again at the $20 level, but they just priced me out for personal assistant, research, and 90% of my token use case. This isn't too surprising to me since they probably have a direct competitor to openclaw et al in the works right now, but until then I am cancelling my subscription and porting my nanoclaw fork with mem0 integration to work with OpenAI instead.Thats not a “That'll teach ‘em” statement, it is just my own cost optimization. I am quite fond of Anthropic's coding models and might still subscribe again at the $20 level, but they just priced me out for personal assistant, research, and 90% of my token use case. Thats not a “That'll teach ‘em” statement, it is just my own cost optimization. I am quite fond of Anthropic's coding models and might still subscribe again at the $20 level, but they just priced me out for personal assistant, research, and 90% of my token use case. a 200 dollar a month customer isn't trying to get around paying for tokens, theyre trying to use the tooling they prefer. opencode is better in a lot of ways.tokens get counted and put against usage limits anyway, unless theyre trying to eat analytics that are CC exclusive they should allow paying customers to consume to the usage limits in however way they want to use the models. They have competition if we don't like it. I don't get more requests for free If you are choosing to pay for a subscription instead, it is because Anthropic is offering those subscriptions at a much better value-per-token. 4 periods of weekly limits, is a monthly limit. Doesn't both count towards my usage limits the same? At its core it's a tragedy of commons situation. Using a third party tool like OpenClaw is augmenting your usage far beyond what was anticipated when the subscription plan was made.Same deal for unlimited storage on drive until people started abusing it. Same deal for unlimited storage on drive until people started abusing it. I didn't set the limits on the plan; change those if it's a problem, not irritate your customer base. The issue is not that it's limited or unlimited, but rather about expected token usage across a user cohort. When you set a usage limit on something like Claude, or a gym, or a tutoring center, you need to do two things at once; set the limit high enough to attract the aspirations of your intended client base ("oh good this gym lets me go every day of the month if I want to"), but priced accurately enough so that you actually turn a profit on the average usage across most users (you ended up going 20 times the first month, but settled into 15 times a month after).If there was suddenly a drug that you could take that would, while you slept, make your body walk to the gym and workout, so that you could max out that usage, the gym would be entitled to adjust either the pricing, the limit, or prohibit going to the gym while on the drug, given that they can't actually sustain all their members going every day.As a correction, I've done some reading and when I said tragedy of the commons, what would fit better is a "congestion externality in a club good". If there was suddenly a drug that you could take that would, while you slept, make your body walk to the gym and workout, so that you could max out that usage, the gym would be entitled to adjust either the pricing, the limit, or prohibit going to the gym while on the drug, given that they can't actually sustain all their members going every day.As a correction, I've done some reading and when I said tragedy of the commons, what would fit better is a "congestion externality in a club good". It's more buying a season pass for Disneyland, then getting told you can't park for free if you're entering the park even though free parking is included with the pass. But 'you can't park even though the ticket includes parking' is not an appropriate analogy because 3rd party use is definitely not intended. They did not 'state one thing' and the 'disallow it'.This is a pretty straight forward case of people using their subscription for 'adjacent' use, and Anthropic being more explicit about it.There's nothing fancy going on here. This is a pretty straight forward case of people using their subscription for 'adjacent' use, and Anthropic being more explicit about it.There's nothing fancy going on here. I did not, in any way, suggest that Anthropic was "pulling the rug" to its users nor that they were entitled to use their tokens using the API with third parties. Full stop.Of course, third-party API usage wasn't intended to be allowed for consuming subscription tokens. This is exactly what my analogy was structured to explain; a Disneyland season pass isn't intended to be used solely for parking. Your analogy missed that last part, which is absolutely crucial to understand.I don't understand how you're making the exact arguments I'm making, then somehow completely misunderstanding what's being said. Of course, third-party API usage wasn't intended to be allowed for consuming subscription tokens. This is exactly what my analogy was structured to explain; a Disneyland season pass isn't intended to be used solely for parking. Your analogy missed that last part, which is absolutely crucial to understand.I don't understand how you're making the exact arguments I'm making, then somehow completely misunderstanding what's being said. in any case Codex is a better SOTA anyways and they let you do this. and if you aren't interested in the best models, Mistral lets you use both Vibe and their API through your vibe subscription api key which is incredible. Many ways, and they're under no obligation to play fair and tell you which way they're using at any given time. You decompile the minified JS code, figure out the protocol, try it from your own code but accidentally mess up a small detail (you didn't realize the nonce has a special suffix). If you share info, they'll probably see it eventually and change their method. They'll probably change it once a month anyway and see who they catch that way (and presumably add a minimum Claude Code version needed to reach their servers).They've got hundreds of super smart coders and one of the most powerful AI models, they can do this all day. So let's say they enforce it by adding an extra nonstandard challenge-response handshake at the beginning of the exchange, which generates a token which they'll expect on all requests going forward. You decompile the minified JS code, figure out the protocol, try it from your own code but accidentally mess up a small detail (you didn't realize the nonce has a special suffix). If you share info, they'll probably see it eventually and change their method. They'll probably change it once a month anyway and see who they catch that way (and presumably add a minimum Claude Code version needed to reach their servers).They've got hundreds of super smart coders and one of the most powerful AI models, they can do this all day. If you share info, they'll probably see it eventually and change their method. They'll probably change it once a month anyway and see who they catch that way (and presumably add a minimum Claude Code version needed to reach their servers).They've got hundreds of super smart coders and one of the most powerful AI models, they can do this all day. If you share info, they'll probably see it eventually and change their method. They'll probably change it once a month anyway and see who they catch that way (and presumably add a minimum Claude Code version needed to reach their servers).They've got hundreds of super smart coders and one of the most powerful AI models, they can do this all day. They've got hundreds of super smart coders and one of the most powerful AI models, they can do this all day. you just need to inspect the network traffic with Claude code and mimic that There are lots of ways they could be doing this. And remember again, if they get you, they don't have to tell you how they got you (so you might not be able to even glean information in return for the $200 you'd be losing).Sure the internet has hundreds of thousands of super smart coders, but the subset who are willing to throw money and credit cards down the drain in order to maintain a circumvention strategy for something like this is pretty low. I'm sure a few people will figure it out, but they won't want to tell anyone lest Anthropic nerf their workaround, so I doubt that exploits of this will become widespread.And if you're Anthropic, that's probably good enough. Sure the internet has hundreds of thousands of super smart coders, but the subset who are willing to throw money and credit cards down the drain in order to maintain a circumvention strategy for something like this is pretty low. I'm sure a few people will figure it out, but they won't want to tell anyone lest Anthropic nerf their workaround, so I doubt that exploits of this will become widespread.And if you're Anthropic, that's probably good enough. There are projects which can do this and hook natively to opencode and even its sdk/api.https://news.ycombinator.com/item?id=47069299#47070204 (I list a project which does this)I really don't know how anthropic can somehow detect something like this. https://news.ycombinator.com/item?id=47069299#47070204 (I list a project which does this)I really don't know how anthropic can somehow detect something like this. I really don't know how anthropic can somehow detect something like this. For example, you can "compact" just portions of the conversation and replace it with a summary. I am not sure how they can detect this. even after this change if you really wanted toBut at this point, to me the question of GP that is that is it even worth it is definitely what I am thinking?I think not. There are better options out there, they mentioned mistral and codex and I think kimi also supports maybe GLM/z.ai as well There are better options out there, they mentioned mistral and codex and I think kimi also supports maybe GLM/z.ai as well There are better options out there, they mentioned mistral and codex and I think kimi also supports maybe GLM/z.ai as well OpenAI will adjust, their investors will not allow money to be lost on ”being nice” forever, not until they're handsomely paid back at least. It suggests to me Anthropic is less concerned with the financial impact of letting subscribers use alternative tools and more concerned with creating lock in to their products and subscriptions. It very well might backfire though, I was not considering alternative models yesterday, but today I am actively exploring other options and considering cancelling my sub. I've been using my subscription primarily through pi recently, so if they aren't interested in me as a customer, pretty much everyone else is. If you are using your own hardware and not profiting directly from Claude's use (as in building a service powered by your subscription). I don't see how this is a problem. I am by no means blowing through my usage (usually <50% weekly with max x5). If you are using your own hardware and not profiting directly from Claude's use (as in building a service powered by your subscription). I don't see how this is a problem. I am by no means blowing through my usage (usually <50% weekly with max x5). Start a new service that is relatively permissive, then gradually restrict APIs and permissions. Finally, start throwing in ads and/or making it more expensive to use. Part of the reason is in the beginning they are trying to get as many users as possible and burning VC money. Then once the honey moon is over, they need to make a profit so they cut back on services, nerf stuff, increase prices and start adding ads. They're losing the exact crowd that they want in their corner because it's the crowd that's far more likely to be making the decisions when companies start pivoting their workflows en-masse. Such artificial "moats" are great for companies but harmful to consumers. Such artificial "moats" are great for companies but harmful to consumers. Trying to prevent competitors from interoperating with the service also may be construed as anticompetitive behaviour.The implementation details of an authentication process do not beget legal privileges to be a monopolist. The implementation details of an authentication process do not beget legal privileges to be a monopolist. https://github.com/agentify-sh/desktopDoes this mean I have to remove claude now and go back to copy & pasting prompts for a subscription I am paying for ? Does this mean I have to remove claude now and go back to copy & pasting prompts for a subscription I am paying for ? I'm just going to accept that my €15 (which with vat becomes €21) is just enough usage to automate some boring tasks. I'm more surprised by people using subscription auth for OpenClaw when its officially not allowed. Things get banned, but that is OK along as they give us weeks or days to prep for alternative solution. Judging from account opening time and comments we can also tell the age group and which camp they are on. - Google reduced AI Studio's free rate limits by 1/10th- Perplexity imposing rate limits, card filing to continue free subscriptions- Now Anthropic as wellThere has been a false narrative that AI will get cheaper and more ubiquitous, but model providers have been stuck in a race for ever more capabilities and performance at higher costs. - Perplexity imposing rate limits, card filing to continue free subscriptions- Now Anthropic as wellThere has been a false narrative that AI will get cheaper and more ubiquitous, but model providers have been stuck in a race for ever more capabilities and performance at higher costs. - Now Anthropic as wellThere has been a false narrative that AI will get cheaper and more ubiquitous, but model providers have been stuck in a race for ever more capabilities and performance at higher costs. There has been a false narrative that AI will get cheaper and more ubiquitous, but model providers have been stuck in a race for ever more capabilities and performance at higher costs. I can get a ridiculous amount of tokens in and out of something like gpt-5.2 via the API for $100.Is this primarily about gas town and friends? Is this primarily about gas town and friends? Instead of using SDKs, this will just shift the third party clients to use ACP to get around it - Claude Code is still under the hood but you're using a different interface.This all seems pretty idiotic on their part - I know why they're trying it but it won't work. There will always be someone working around it. There will always be someone working around it. So it makes sense to offer simple flat pricing for first party apps, and usage priced apis for other usage. We're still collectively figuring out the best way to use models in software, this TOS change kills innovation. But if you're doing something very basic, you might be able to slop together a tool that does local inferencing based on a small, local model instead, alleviating the need to call Claude entirely. Can't this restriction for the time being be bypassed via -p command line flag? Anthropic = goodGoogle = evilThat's pretty much HN crowd logic to be honest Google = evilThat's pretty much HN crowd logic to be honest That's pretty much HN crowd logic to be honest So, I guess it's time to look into OpenAI Codex. Opencode with CC underneath using Gigacode?OpenAI codex is also another viable path for what its worth.I think the best model to my liking open source is kimi k2.5, so maybe you can run that?Qwen is releasing some new models so I assume keep an eye on those and maybe some model can fit your use case as well? OpenAI codex is also another viable path for what its worth.I think the best model to my liking open source is kimi k2.5, so maybe you can run that?Qwen is releasing some new models so I assume keep an eye on those and maybe some model can fit your use case as well? I think the best model to my liking open source is kimi k2.5, so maybe you can run that?Qwen is releasing some new models so I assume keep an eye on those and maybe some model can fit your use case as well? Qwen is releasing some new models so I assume keep an eye on those and maybe some model can fit your use case as well? It's a little bit sleazy as a business model to try to wedge one's self between Claude and its users.OpenAI acquiring OpenClaw gives me bad vibes. It doesn't seem organic.I definitely feel much more aligned with Anthropic as a company. What they do seems more focused, meritocratic, organic and genuine.OpenAI essentially appropriated all their current IP from the people... They basically gutted the non-profit and stole its IP. Then they used media spin to make it sound like they appropriated it from Elon because Elon donated a few millions... I mean let's not even mention Suchir Balaji, the OpenAI researcher who supposedly "committed suicide" after trying to warn everyone about the stolen IP.OpenAI is clearly trying to slander Anthropic, trying to present themselves as the good guys after their OpenClaw acquisition and really rubbing it in all over HN... Over which they have much influence. It doesn't seem organic.I definitely feel much more aligned with Anthropic as a company. What they do seems more focused, meritocratic, organic and genuine.OpenAI essentially appropriated all their current IP from the people... They basically gutted the non-profit and stole its IP. Then they used media spin to make it sound like they appropriated it from Elon because Elon donated a few millions... I mean let's not even mention Suchir Balaji, the OpenAI researcher who supposedly "committed suicide" after trying to warn everyone about the stolen IP.OpenAI is clearly trying to slander Anthropic, trying to present themselves as the good guys after their OpenClaw acquisition and really rubbing it in all over HN... Over which they have much influence. I definitely feel much more aligned with Anthropic as a company. What they do seems more focused, meritocratic, organic and genuine.OpenAI essentially appropriated all their current IP from the people... They basically gutted the non-profit and stole its IP. Then they used media spin to make it sound like they appropriated it from Elon because Elon donated a few millions... I mean let's not even mention Suchir Balaji, the OpenAI researcher who supposedly "committed suicide" after trying to warn everyone about the stolen IP.OpenAI is clearly trying to slander Anthropic, trying to present themselves as the good guys after their OpenClaw acquisition and really rubbing it in all over HN... Over which they have much influence. OpenAI essentially appropriated all their current IP from the people... They basically gutted the non-profit and stole its IP. Then they used media spin to make it sound like they appropriated it from Elon because Elon donated a few millions... I mean let's not even mention Suchir Balaji, the OpenAI researcher who supposedly "committed suicide" after trying to warn everyone about the stolen IP.OpenAI is clearly trying to slander Anthropic, trying to present themselves as the good guys after their OpenClaw acquisition and really rubbing it in all over HN... Over which they have much influence. OpenAI is clearly trying to slander Anthropic, trying to present themselves as the good guys after their OpenClaw acquisition and really rubbing it in all over HN... Over which they have much influence. > The companies that will win long-term are the ones building open protocols and letting users bring their own model.These seem contradictory. The future I see is open source harnesses talking to commodity models. The future I see is open source harnesses talking to commodity models. Perhaps there's enough overlap among the low hanging fruit that you can initially sell a harness that makes both genomics researchers and urban planners happy... but pretty quickly you're going to need to be the right kind of specialist to build an effective harness for it. Try it: write your own harness with (bash, read, write, edit) ... it's trivial to get a 99% version of (pick your favorite harness) -- minus the bells and whistles.The "magic of the harness" comes from the fun auxiliary orchestration stuff - hard engineering for sure! It turns out that "most of the bell and whistles" could amount to instructing models how to use tools like tmux Unfortunately neither political party can get all of the above. People self-hosting don't look at their electricity bill. So, which two parties could they be referring to? All that to say, don't let the naysayers get you down. I bought my Mac Mini last week and have been really happy with it as an isolated environment. The always-on nature of OpenClaw means that it's nice to be able to restart my personal laptop or do gaming or whatever else I want and I'm not fighting for GPU resources in the background. It was animating a blue dot with my real time position at the same time.Don't get me started about the new iOS 26 notification and messaging filters. Those are causing real harm multiple times a day. Don't get me started about the new iOS 26 notification and messaging filters. Those are causing real harm multiple times a day. The markets value recurring subscription revenue at something like 10x “one-off” revenue, Anthropic is leaving a lot of enterprise value on the table with this approach.In practice this approach forces AI apps to pay Anthropic for tokens, and then bill their customers a subscription. And many categories of free app are simply excluded, which could in aggregate drive a lot more demand for subscriptions.If Anthropic is worried about quota, seems they could set lower caps for third-party subscription usage? (Maybe this is purely about displacing other IDE products, rather than a broader market play.) In practice this approach forces AI apps to pay Anthropic for tokens, and then bill their customers a subscription. And many categories of free app are simply excluded, which could in aggregate drive a lot more demand for subscriptions.If Anthropic is worried about quota, seems they could set lower caps for third-party subscription usage? (Maybe this is purely about displacing other IDE products, rather than a broader market play.) If Anthropic is worried about quota, seems they could set lower caps for third-party subscription usage? (Maybe this is purely about displacing other IDE products, rather than a broader market play.) (Maybe this is purely about displacing other IDE products, rather than a broader market play.) Allows them to optimize their clients and use private APIs for exclusive features etc. and there's really no reason to bootstrap other wannabe AI companies who just stick a facade experience in front of Anthropic's paying customer. Tell me if they are eating losses or not. I'm not making a point, actually do it and let me know. That means I'm subsidizing them (paying 3-4 times what it would cost with the API)! Over 9 days I would have spent roughly $63 dollars on Codex with 11.5M input tokens plus 141M cached input tokens and 1.3M output tokens.That roughly mirrors the $100-200/wk in API spending that drove me to the subscription. This is a 30/d window and I only used it for 9 days, $63 worth. OpenAI kept the other $137.It makes sense though for heavy use. This is a 30/d window and I only used it for 9 days, $63 worth. OpenAI kept the other $137.It makes sense though for heavy use. This is a 30/d window and I only used it for 9 days, $63 worth. OpenAI kept the other $137.It makes sense though for heavy use. DeepSeek has the tendency to think... a lot!. Without a good harness I can't evaluate it well; time will tell.OpenAI doesn't; it's embedded into the price, I think.Cheap = we can run 10x the workloads, bigger imagination = innovation. Maybe 10 dumb agents in a loop can beat 1 Opus? Maybe 10 dumb agents in a loop can beat 1 Opus? Cheap = we can run 10x the workloads, bigger imagination = innovation. Maybe 10 dumb agents in a loop can beat 1 Opus? I don't think Anthropic has any desire to be some B2C platform, they want high paying reliable customers (B2B, Enterprise). Cloud goes on the books as recurring revenue, not one-off; even though it's in principle elastic, in practice if I pay for a VM today I'll usually pay for one tomorrow. (I don't have the numbers but the vast majority of cloud revenue is also going to be pre-committed long-term contracts from enterprises. )> I don't think Anthropic has any desire to be some B2C platformThis is the best line of argument I can see. (I don't have the numbers but the vast majority of cloud revenue is also going to be pre-committed long-term contracts from enterprises. )> I don't think Anthropic has any desire to be some B2C platformThis is the best line of argument I can see. > I don't think Anthropic has any desire to be some B2C platformThis is the best line of argument I can see. This is the best line of argument I can see.

Mark Zuckerberg walked somewhat stiffly into a courtroom of the Superior Court of Los Angeles County at 9 am on the dot on Wednesday, having first been escorted into the building by a security detail that included two Department of Homeland Security officers. The venue, overseen by Judge Carolyn Kuhl, was bursting with spectators and media, many crammed elbow-to-elbow on benches, all there to witness the Meta CEO testify before a jury about allegations that his company's products pose serious risks to younger users. They allege that her compulsive use of those platforms accounts at an extremely young age caused severe psychological damage. These are the cases selected as representative of a much larger pool of litigants—some 1,600 in total—who have filed suits against the same brands on similar grounds, alleging that their children fell victim to depression, dysmorphia, and suicide after they got hooked on apps that absorbed their attention. The families of some of these children were among those vying for a chance to finally see Zuckerberg in the hot seat on Wednesday. Lanier picked apart claims Zuckerberg had made under oath in January 2024 during congressional testimony about online child safety. Elsewhere, Lanier sought to establish Zuckerberg's ultimate decisionmaking authority over Meta and quoted a remark he made in an interview with Joe Rogan last year: “Because I control our company, I have the benefit of not having to convince the board not to fire me,” he'd told the podcast host. Zuckerberg insisted in court that this was merely a “simplified” version of the truth. Over the course of his testimony, Zuckerberg was oddly evasive about even insignificant details and basic definitions. He was not quite prepared to confirm that his relevant congressional testimony had taken place on January 31, 2024, for example, and was hesitant to agree with Lanier's proposition that when something is “addictive,” people “do it more.” He ultimately hedged: “Maybe in the near term.” Zuckerberg also preferred to acknowledge any past comment, in probabilistic fashion, by saying, “It sounds like something I would have said.” Likewise, when asked whether internal documents appeared to suggest Meta's interest in maximizing “total teen time spent” on their apps, he often replied, “That's what the document says.” Zuckerberg repeatedly fell back on accusing Lanier of "mischaracterizing" his previous statements. When it came to emails, Zuckerberg typically objected based on how old the message was or his lack of familiarity with the Meta employees involved. And Zuckerberg never failed to point out when he wasn't actually on an email thread entered as evidence. Perhaps anticipating these detached and repetitive talking points from Zuckerberg—who claimed over and over that any increased engagement from a user on Facebook or Instagram merely reflected the “value” of those apps—Lanier early on suggested that the CEO has been coached to address these issues. Lanier went on to present Meta documents outlining communication strategies for Zuckerberg, describing his team as “telling you what kind of answers to give,” including in a context such as testifying under oath. “I'm not sure what you're trying to imply,” Zuckerberg said. In an even more, well, “meta” moment after the court had returned from lunch, Kuhl struck a stern tone by warning all in the room that anyone wearing “glasses that record”—such as the AI-equipped Oakley and Ray-Ban glasses sold by Meta for up to $499—had to remove them while attending the proceedings, where both video and audio recordings are prohibited. 's suit and the others to follow are novel in their sidestepping of Section 230, a law that has protected tech companies from liability for content created by users on their platforms. When Lanier presented evidence that Meta teams were working on increasing the minutes users spent on their platforms each day, Zuckerberg countered that the company had long ago moved on from those objectives, or that those numbers were not even “goals” per se, just metrics of competitiveness within the industry. When Lanier questioned if Meta was merely hiding behind an age-limit policy that was “unenforced” and maybe “unenforceable,” per an email from Nick Clegg, Meta's former president of global affairs, Zuckerberg calmly deflected with a narrative about people circumventing their safeguards despite continual improvements on that front. Lanier, though, could always return to K.G.M., who he said had signed up for Instagram at the age of 9, some five years before the app started asking users for their birthday in 2019. While Zuckerberg could more or less brush off internal data on, say, the need to convert tweens into loyal teen users, or Meta's apparent rejection of the alarming expert analysis they had commissioned on the risks of Instagram's “beauty filters,” he didn't have a prepackaged response to Lanier's grand finale: a billboard-sized tarp, which took up half the width of the courtroom and required seven people to hold, of hundreds of posts from K.G.M. When Lanier had finished and Schmidt was given the chance to set Zuckerberg up for an alternate vision of Meta as a utopia of connection and free expression, the founder quickly gained his stride again. “I wanted people to have a good experience with it,” he said of the company's platforms. Then, a moment later: “People shift their time naturally according to what they find valuable.” A wave of unexplained bot traffic is sweeping the web Big Story: The women training for pregnancy like it's a marathon Listen: Silicon Valley tech workers are trying to stop ICE The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.