Jason de Koning 🇨🇦

Wait...you're "testing" if a small number of poor bastards who signed up to give you $$ care if they don't get the same service that every one else gets for the same price? This is supposed to make people feel better? What are you guys doing? You should delete this tweet and apologize.

For clarity, we're running a small test on ~2% of new prosumer signups. Existing Pro and Max subscribers aren't affected.

Mythos is not a bad name for a model but it would be better if Anthropic switched to using famous Claudes. Monet, Debussy etc. The final model that achieves AGI would obviously be Van Damme

@joshgonsalves_ Same here. It's too bad, it was really liking its output

Oh, so Claude Design has it's own usage limit outside of everything else? And of course, already hit it... So now I can't use it until NEXT FRIDAY? OK...

@helloitsaustin It's a great tool, but I used up almost my entire weekly limit casually working on a single slide deck (less than an hour on the 20X max plan) :(

we just launched one of the coolest features for marketers today: Claude Design 🎨 I've posted a lot about building marketing tools without being technical, and this is the visual side of that. I described a landing page editor and Claude built me one. here's an example of my actual workflow with it 👇

@OpenAI @Amgen @moderna_tx @AllenInstitute @thermofisher This looks exciting, but it's really disappointing to see this as a closed research preview. Why not make this freely available to academic scientists already paying for pro/max subscriptions?

Our Life Sciences model series is available as a research preview starting today for qualified customers including @Amgen, @moderna_tx, the @AllenInstitute, and @thermofisher Scientific through ChatGPT, Codex, and the API. openai.com/index/introduc…

Introducing GPT-Rosalind, our frontier reasoning model built to support research across biology, drug discovery, and translational medicine.

@MilkRoadAI Two Nobel-recognized researchers? LeCun has not won a Nobel prize (though he shared a Turing award - no shade intended)

The CEO of Google DeepMind just went on record saying he disagrees with one of the most respected AI researchers in the world. Demis Hassabis, the man behind AlphaFold, AlphaGo, and Google's entire AI operation publicly pushed back against Yann LeCun's claim that large language models are a dead end for artificial intelligence. LeCun, who left Meta earlier this year to start his own AI lab, has been saying for years that LLMs cannot reason, cannot plan, and will never get us to human-level intelligence. Hassabis disagrees, and he said so directly. His position is that scaling laws are still working, foundation models are still getting more capable, and whatever AGI ends up looking like, LLMs will be a central part of it, not something that gets replaced. He does say there is roughly a 50/50 chance that one or two additional breakthroughs will be needed beyond scaling alone, things like better memory, long-term planning, and world models. But the core disagreement with LeCun is clear, Hassabis believes the current architecture is sound and the current path leads somewhere real. Two Nobel-recognized researchers, two founding figures of modern AI, now publicly on opposite sides of the most important technical question in the industry.

@chatgpt21 Elon says many things. Whether the truth is among them is unclear.

Elon Musk says it will take till May for Grok to be close to Opus 4.6 & June to “match or exceed standards” (see red dots) - that’s where Grok & 4.6 Opus line up today on AAIA. But xAI isn’t competing with Opus 4.6 they’re competing with Mythos. And if you do even rough indexing math, Mythos likely lands 60–61+. At that point I’ll say it again: I don’t know if it’s even possible to catch up under the current paradigm. If Anthropic employees are each running 4 instances of Claude Mythos, they’re effectively out-building you by 30–50x. If you’re not already fully leveraging systems like Claude Code or Codex the way Meta is starting to, you fall behind fast. This is why Anthropic removed xAI from Cursor which sent XAI to the shadow realm - (I’ll make another post on this later) And if Mythos is already working on the next iteration, then by the time Grok “matches” Opus 4.6 Claude is already one model ahead, maybe two (Mythos 2.0). you just simply can’t chase a moving target that’s accelerating faster than you. (Unless you run a different race) xAI probably needs to focus on a different research paradigm to win this race because in this one, you can’t mathematically catch Anthropic.

New preprint! @JShendure @CXchengxiangQIU Can we learn regulatory grammars of human cell types we'll never be able to profile - by training on mouse development and transferring across 241 mammalian genomes?biorxiv.org/content/10.648…

@mgdurrant Yes! When?

Excited for scientists to start using Mythos-class models for their work!

@DarioAmodei Bravo! It would be great if you could also keep academic scientists working in the public interest in mind for priority access. I promise, we'll put it to good use.

The dangers of getting this wrong are obvious, but if we get it right, there is a real opportunity to create a fundamentally more secure internet and world than we had before the advent of AI-powered cyber capabilities.

I’m proud that so many of the world’s leading companies have joined us for Project Glasswing to confront the cyber threat posed by increasingly capable AI systems head-on. x.com/AnthropicAI/st…

@alexalbert__ Wow. ...I get the emphasis on cyber, but I hope @AnthropicAI will also consider academic scientists as a priority group for expanding (affordable) access ASAP.

We released Claude Opus 4.6 just two months ago. Today we're sharing some info on our new model, Claude Mythos Preview.

@Cappy_Nate I’m a computational biologist and am interested! I’m working on Bio x AI, health, genomics, etc.

Started an Actual Canadian Builders group chat. -Hardware -Software -Space -Defence -Web3 -Health -AI -Energy -Manufacturing You name it. But it's a Canadian only zone. 🇨🇦 If you want in, drop your name in the comments and ping anyone who should be there! LFG! 💪🇨🇦🏗️🚀

@bensig ..."Leeloo Dallas multipass", Milla??

My friend Milla Jovovich and I spent months creating an AI memory system with Claude. It just posted a perfect score on the standard benchmark - beating every product in the space, free or paid. It's called MemPalace, and it works nothing like anything else out there. Instead of sending your data to a background agent in the cloud, it mines your conversations locally and organizes them into a palace - a structured architecture with wings, halls, and rooms that mirrors how human memory actually works. Here is what that gets you: → Your AI knows who you are before you type a single word - family, projects, preferences, loaded in ~120 tokens → Palace architecture organizes memories by domain and type - not a flat list of facts, a navigable structure → Semantic search across months of conversations finds the answer in position 1 or 2 → AAAK compression fits your entire life context into 120 tokens - 30x lossless compression any LLM reads natively → Contradiction detection catches wrong names, wrong pronouns, wrong ages before you ever see them The benchmarks: 100% recall on LongMemEval — first perfect score ever recorded. 500/500 questions. Every question type at 100%. 92.9% on ConvoMem — more than 2x Mem0's score. 100% on LoCoMo — every multi-hop reasoning category, including temporal inference which stumps most systems. No API key. No cloud. No subscription. One dependency. Runs on your machine. Your memories never leave. MIT License. 100% Open Source. github.com/milla-jovovich…

@bcherny Can you pls clarify if this a policy change since the Jan crackdown on OAuth mis/use with OpenCode, OpenClaw, etc., or is this just a notice that the previous policy will now be strictly enforced? Are ocassional cron-based callouts to `claude -p` still kosher?

We want to be intentional in managing our growth to continue to serve our customers sustainably long-term. This change is a step toward that.

Starting tomorrow at 12pm PT, Claude subscriptions will no longer cover usage on third-party tools like OpenClaw. You can still use these tools with your Claude login via extra usage bundles (now available at a discount), or with a Claude API key.

@thsottiaux @OpenAICodexCli In the last 24 hours i've been intermittently getting garbled text from Codex using gpt-5.4 xhigh. I've never seen this before. Are you guys on this? e.g.: "The current rer且 the ` will likely now run to completion; meanwhile I'm waiting on its artifact summary so below is notrə. Irotma: the last interrackh help【accd090000 No ops / no changes. obviator overlapping thrust just contd0 C no更. entered001 -; existing currently yes."

@peterottsjo 👀 nature.com/articles/s4158…

How much of biological AI actually works? The answer, according to a new Nature Biotechnology review, is damning. I took a look at it. Here's what I found.🧵

$100 in Codex credits for college student in U.S. and Canada:

@tszzl You must be trolling

the dune movies were doomed from the start to be good and not great due to the casting of chalamet as paul. he does not have the gravitas for a child-god and is much better suited for kind of silly coming of age movies

@amorriscode @tannerlinsley Need any scientist beta testers? I would love to give it a trial

@tannerlinsley let me know if you want in on the beta for the new CC Desktop UI/UX

Ghostty was fun, but time for something else. I still love opencode, too but with CC plans dead on it… I’m feeling lost. Full GUI? T3 Code? Opencode GUI? Warp? Back to cursor? Try CC again? Raw Codex? My 🧠 hurts and I just need to keep shipping.

@MatthewBerman @nvidia @Dell Congrats!! Want. (For science purposes.)

.@nvidia hand delivered a pre-production unit of the @Dell Pro Max with GB300 to my house. 100lbs beast with 750GB+ of unified memory to power the best open-source models in the world. What should I test first?