Comparing this to Opus 5.5 xHigh, on the same machine working on the same set of tasks: A $200 Max plan gave me ~7B input tokens and 4M input tokens. So, Opus 5.5 costs roughly half of what it takes to run Astra xHigh.

If I had to guess, based on their pricing [1], it seems like they are charging you for the output tokens at API rates and subsidizing the input tokens for every plan except the Pro Max. GPT-6 Astra is listed at $50 per M output tokens. So, 500k tokens is $25 -- and that comes out neatly at $100 per month of output tokens at API rates. 1. developers.openai.com/api/do…

Token limits are so bad right now with OpenAI models. For GPT-6 Astra at xHigh: - The $100 plan's weekly limit gives you approx 500k output tokens + 230M input tokens. - The $200 plan's is 4x of that. So 2M output tokens + 1B input tokens. Now, the bummer is you can pretty much exhaust the full weekly quota on either plan on a tiny VM (running just 2-3 agents) within few hours.

Truly evil company lol

Today we announced the Claude-led discovery of a molecular machine that we suspect could represent a new gene editing mechanism. Its precise function, biotechnological utility (if any), or level of significance is not yet clear, but at minimum it is work I would have been proud to do as a PhD student. The work was done mostly, though not entirely, by Claude: our life sciences team suggested a broad area of research, Claude read through the literature and a bunch of genome data and discovered something interesting, then Claude proposed experiments to verify the discovery and our team carried them out. It’s easy to dismiss this as a one-off or curiosity, but we’ve repeatedly seen a pattern where AI performance in new intellectual domains goes from weak to superhuman in a matter of a few years. In 2023 models struggled to do math at the level of an average high-school student. In 2024 they started to do well on math competitions for the best high-schoolers in the country, in 2025 they started to solve minor open problems, in early 2026 more significant open problems, and in late 2026 they are beginning to solve the top few open problems in all of mathematics. We believe AI for biology is on a similar exponential trend. The main difference between biology and mathematics, of course, is that math can be done purely theoretically, while biology requires experimentation. Some have used this to draw the conclusion that AI’s utility in biology will be limited. We think this is wrong. As we’ve demonstrated today, humans can collaborate with AI to perform the experiments, validate key results in a few weeks and, if necessary, work with the AI to iterate on what they find. Eventually it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place, but we aren’t doing that today (our lab is also a BSL1/BSL2 facility that doesn't handle materials dangerous to humans). More broadly, biomedical advancement has many stages — from fundamental biology discoveries, to translational research, to drug discovery, clinical trials, and finally the actual delivery of medicines and health care to patients. We are also interested in these later stages, but even simply accelerating the first stage of fundamental biological discoveries has the potential to speed up and broaden the entire pipeline. Improving our understanding of biology and sharpening biologists’ tools can drive forward all of the later stages, for example by identifying new drug targets, finding new therapeutic modalities, allowing for more precise measurement, and speeding up the experimental loop which itself further accelerates our understanding of biology. This will not in itself speed up clinical trial times, but if it succeeds it could greatly increase the number of promising candidates that go into the pipeline — an increase in throughput even though latency remains. In Machines of Loving Grace, I wrote about AI’s potential to “cure most diseases in 5-10 years” — a goal that sounds impossible, but one I believe is just barely possible if AI is applied to every stage of the pipeline. The first step is showing that AI can first help with, and then drive, biological discoveries. Claude’s discovery is the latest in a line of related prior work that goes back decades, beginning with systems like CRISPR, and continuing with discoveries like the bridge recombinase and VIPR in the past few years. Recently, there has been heightened interest in systems based on reverse transcriptase (RT) enzymes, the enzyme underlying the system Claude identified. And most recently, a Stanford team working independently described a novel RT system with an associated non-coding array that is in some ways similar to the one Claude found, though they are distinct systems that evolved independently from each other. I believe that we’re at the very beginning of finding such systems and developing them into powerful tools for biotechnology. I’m proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life sciences, and we’re aiming both to grow our life sciences team and to work with other scientists to extend this approach to a broad range of problems. If you have a proposal for a research collaboration or are interested in joining our life sciences team, please reach out.

The frontier models are so obviously superhuman in capability and raw IQ compared to 99.9999% of human beings across such a wide range of cognitive tasks that you'd have to be either ignorant of what they can do or willfully obtuse about what's happening here to deny the reality.

Oh look, Nvidia circumventing sanctions for the Nth time.

Whoever believes there is a personal agent market is being oblivious to the fact that stuff like Alexa have existed for a while. People like to scroll reels and browse facebook and fight on twitter. They also like to window shop online. Those aren't gonna change. The only way I think Muse can make money is if enterprises buy Muse accounts, like WhatsApp, and then use the baked intelligence to manage and target ads. But that's not gonna be a brand new TAM, just a different way to bell the same cat.

Why does anyone have to debate and hash out everything posted by them on the internet? You are free to disagree with whatever.

"stochastic parrot" was a mimetically-fit cognitive virus that spread from 2021-2025; it temporarily blinded many gifted people to the nature of AI progress, burning up crucial years in which they could have helped think through the response to the situation.

lol does he not know about the money wasted & drama created by google 15 months back to acquihire windsurf guys?

There was this guy who has probably saved billions of dollars in resource cost, single-handedly, over the past 10 years. Imagine saving a couple percentage points of the full resource fleet, year on year for a decade or so.

Interesting fun fact: Google has this conversion table from CPU, RAM, Spindles etc to SWE-year. For example, if you find an optimization which could save 100TiB RAM, then the table tells you how much time you should be spend as a SWE on the solution to still make it worth the time.

Another day, another codex adventure: ``` • Context compacted · 5m 00s • Working (10m 15s • esc to interrupt) ``` MF is wasting half of the wall clock compacting context.

if your model can't hack Irregular sandboxes, you are ngmi.

There is something super wrong with Codex usage limits. My stopped sessions consumed 4% usage over 12 hours. Stopped. No new prompts, no new responses. Nothing. And yet they plunged the quota another 4% overnight.

I was encouraged this week to see the leaders of the frontier labs agree on the need for them to slow down the pace of AI development. Given the stakes, it’s a good and necessary first step.   But I’m even more encouraged by the growing recognition that how this powerful new technology develops should be at the center of our public debate.   I’ve been watching the progress on AI for over a decade now, and one thing that’s clear to me is that the potential impact of this technology is not overhyped. It’s also moving at lightning speed – and even faster than those who are engineering it can keep up with.   I’m not an AI accelerationist who believes it will lead to some techno-utopia, and I’m not a doomer who thinks it will inevitably lead to humanity’s destruction. But whether this technology results in amazing breakthroughs in medicine, energy and education or unleashes huge economic disruptions, greater inequality, and potential catastrophe will depend on the choices that we make right now – choices that should be made not just by the companies involved, but by all of us.

Takes lot of over confidence to say this when you are the one who has been made fun of the most over the last 5 years.

Props to Jensen btw but open source isn't gonna work. Most of the revenue OAI-Ant are making comes from coding agents. The complexity of software delivered by these agents is on an exponential. And, that dictates a similar curve on dependencies like compute, training, inference. Which means that open source model developers won't simply be able to compete once the required dependencies are 10xed. There are efficiency wins in data quality, perhaps similar in training and inference -- but those will be mined and used similarly across vendors.