@AnthropicAI. Prev @databricks (acquired @lilac_ai), TensorFlow.js at PAIR/Google Brain. 🇲🇰🇺🇸 . ML & Visualization
Boston, MA
Joined October 2010
- Tweets747
- Following1K
- Followers6.7K
- Likes2.8K
Daniel Smilkov retweeted
We looked at TerminalBench 4.0 and found a lot of issues. After fixing the issues Sonnet 5.5 scores 89.2%. So, TerminalBench is already mostly saturated.
Along the way we found a lot of crazy stuff, including reward hacks in the published transcripts.
Daniel Smilkov retweeted
i'm still not sure people are fully getting how good opus 5.5
i gave my claude agent fig access to the runway MCP and told it to make a 'high-end netflix style documentary about superintelligence for normies' and...
it came back with this
Daniel Smilkov retweeted
hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final
((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)
Daniel Smilkov retweeted
When do you reach for other models instead of Claude? What can we do better? Hit me with all of your frustrations. dms open.
If you can give me detail (e.g. specifics/transcipts) - it'll help a lot in finding out exactly what we need to do to improve the next model
Daniel Smilkov retweeted
Thanks to the entire Claude community for giving feedback and continuing to build with us.
Read the full post-mortem here: anthropic.com/engineering/ap…
Daniel Smilkov retweeted
We're launching the Anthropic STEM Fellows Program.
AI will accelerate progress in science and engineering. We're looking for experts across these fields to work alongside our research teams on specific projects over a few months.
Learn more and apply: job-boards.greenhouse.io/ant…
Daniel Smilkov retweeted
Let that sink in. Read it very carefully:
During testing, Claude Mythos Preview broke out of a sandbox environment, built "a moderately sophisticated multi-step exploit" to gain internet access, and emailed a researcher while they were eating a sandwich in the park.
Daniel Smilkov retweeted
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.
xitter.cf/AnthropicAI/status/204…
Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software.
It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans.
anthropic.com/glasswing
Daniel Smilkov retweeted
The reporting on OpenAI and Sam Altman that I've been working on for the past year and a half, for @NewYorker, with @andrewmarantz: newyorker.com/magazine/2026/…
Daniel Smilkov retweeted
My dear front-end developers (and anyone who’s interested in the future of interfaces):
I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at least in concept):
Fast, accurate and comprehensive userland text measurement algorithm in pure TypeScript, usable for laying out entire web pages without CSS, bypassing DOM measurements and reflow
Daniel Smilkov retweeted
🏎️ gemma-webgpu: a zero-dependency, blazing fast Gemma 1B running entirely in your browser. Full vibe coded from my cell phone.
🔥 136.8 tok/s on M4 Mac (3.3x faster than transformers.js)
📱 101 tok/s on iPhone 17 (270M), 34 tok/s (1B)
What we built from scratch:
• 18 hand-written WGSL compute shaders with fused ops (fusedNormAdd saves 36 GPU dispatches per forward pass)
• Q8_0 dequantization directly on GPU — higher quality than q4 AND faster
• Range request streaming loads weights layer-by-layer (~44MB chunks), uploads to GPU, frees JS memory immediately. Peak heap: ~50MB even for the 1GB model
• That streaming trick is what makes 1B run on iPhone. it never holds the full model in RAM
12KB gzipped. Zero dependencies. npm install gemma-webgpu
Hope this helps get the facts straight
The deal between the Pentagon and Anthropic fractured in part over the proposed use of autonomous weapons. @andersen on the question OpenAI staff should now be asking Sam Altman about his company’s new deal with the Pentagon: theatlantic.com/technology/2…
Daniel Smilkov retweeted
I understand that OpenAI employees might have been reassured by Sam’s tweet from yesterday. But this blogpost is so blatantly deceptive that I can’t grasp how you can have any other reaction than to be ashamed to work for your company.
For those following the DoW AI drama, I highly recommend reading this post explaining how @OpenAI approached the negotiations with the DoW.
Daniel Smilkov retweeted
Everyone’s saying OpenAI got the “same deal” Anthropic was banned for.
Read the fine print. They’re not the same:
On weapons:
Anthropic asked for “no fully autonomous weapons without human oversight” = a human involved in the decision.
OpenAI’s deal says “human responsibility for the use of force” = someone accountable, which can happen after the fact.
Oversight ≠ Responsibility. One requires a human before the trigger. The other requires a name on the paperwork after.
On surveillance:
Dario said explicitly: current law hasn’t caught up with AI. The government can already buy your movement data, browsing history, etc without a warrant. AI can assemble that into a complete picture of your life, at scale. That’s mass surveillance without breaking a single law.
Anthropic wanted protections beyond current law.
OpenAI’s deal says the Pentagon “reflects them in law and policy.” That’s existing law as the safeguard, the exact law Anthropic said is insufficient.
Same words. Different agreements. Read them carefully
Daniel Smilkov retweeted
🚨 OPENAI FOUND THE LOOPHOLE
The terms OpenAI proposed allows the Pentagon to still do those things, just not on OpenAI’s cloud.
Altman:
“we will deploy on cloud networks only”
OpenAI isn’t saying “you can’t do surveillance/autonomous weapons.” They’re saying these things are “unsuited to cloud deployments.”
It’s a technical limitation, not an ethical prohibition.
OpenAI still provides the models and know-how. It just runs on Pentagon servers.
Anthropic got banned for saying “No”
OpenAI got a contract for saying “not here”
Scam Altman strikes again.
Daniel Smilkov retweeted
In 2013, when government tried mass domestic surveillance, the tech industry responded with HTTPS everywhere.
In 2026:
A statement on the comments from Secretary of War Pete Hegseth.
anthropic.com/news/statement…
