一个战略官用脚投票的故事
I am excited to share that I have joined Anthropic. I am working to ensure that Claude practices and values epistemic rigor. This continues my long-standing interest in research methodology: how we learn what we know, and how we quantify what we do not.
I am grateful to @demishassabis, who hired me at Google DeepMind, to @sundarpichai, and to the many wonderful people I got to know there.
21
走出自己的舒适区在这场变革中尤为重要,当然保持健康是前提
Subtle reminder that you're supposed to feel overwhelmed.
It means you're doing something new. You're taking a new path. You're learning. Growing. You branched into the unknown. It will be frustrating. You will feel lost. It will be painful, but staying the same is even worse.
16
Anthropic 也越界了,这点无可争辩。他们摧毁了数百万本书籍——人类掌握的知识——并据为己有。我们应该永不原谅他们
A big Chinese lab reached out to me asking for help distilling Fable into their model. I said no.
I told them to use the millions of books that have no copyright instead. Offered to help pick them and to help train the model a different way. They only wanted distillation, which tells you a lot about the mindset there.
I've worked deep in this field and distillation is real. I wasn't convinced at one point. What I saw changed my mind completely.
I'm the biggest supporter of open source. I'm for releasing weights and for the freedom to research and build, because humanity only grows when knowledge gets shared. But let's be logical.
Nobody will convince me that labs shipping a model every month or less, each one a big leap, now can't ship a better one after three months.
Kimi K3.1 Pro isn't a big leap over Kimi K3 and it's been about 100 days.
DeepSeek still can't put out a respectable Pro model.
Zai used to release their strongest model in under two months. It's been three with nothing.
All I hear about Qwen 4 Max is that it's a letdown.
Kimi K3.1 Pro and DeepSeek Flash both show traces of Astra distillation. The leak showed Moonshot used it, and so did DeepSeek.
If this isn't the effect of Anthropic blocking distillation then what is it?
Chinese labs need to get serious about using the talented researchers they have and give their models an identity. Right now their models at their best are a cheaper alternative. I want them out of that box. I want to see what makes them different.
Anthropic has crossed lines too, no argument there. They destroyed millions of books, human knowledge that was in people's hands, and kept it for themselves. I'll never forgive them for that.
But Chinese labs rely on distillation more than they rely on better data.
I didn't work with them. Transparency and honesty matter to me, like you've known since I started here.
29
可以了,真的可以了
我去!Fable 5.2 只用了一下午、45 亿 token 就给我做出了这样一个 3D 动作游戏!
包括模型、场景设计、动作设计、优化都是它搞的,这一切都被融合在了一个 30 兆的 HTML 文件里边。
总共就花了一下午,它这个游戏制作能力太恐怖了
可以在这里试试:op7418.github.io/guizang-gam…
20
以前这种体验可是收费的
I wanted to test Fable 5.2's quality with one-shot prompts without human feedback.
Everything in this game is from scratch: models, sounds, textures. ~$550 at API pricing, ~2.7M output and 445M input tokens, heavily cached. Took 8 hours!
Play it here: claude.ai/artifact/BGY1SztrA…
1
25


















