Bo Lu

I write code and take pictures

[Bo Lu @blu3ness]

Most people see Bayesian inference as “updating beliefs.”

That’s superficial.

The real power is factorization: models decompose into likelihood blocks that can be plugged together.

A thread.

[Bo Lu @blu3ness]

@ZacharyZhang @dhh Thank you for open sourcing this and sharing your excellent library with the world

[Shawn McAllister @eas4ai]

Holy sheet!!! 🤯

[Yoshua Bengio @Yoshua_Bengio]

Over the past few days, I've taken the time to summarize my thoughts on the recent incidents involving agents’ misaligned behavior. We don't know with certainty what comes next, but we know where these issues originate, and this can help us plan the path forward.

Please feel free to ask your questions in the replies, and I’ll try to answer some of them in the coming weeks.

https://t.co/bvncjT0y0h

[DAN KOE @thedankoe]

https://t.co/7l7Jef99QZ

[Hux @huxlab]

今晚别刷Netflix了。

去看这个SpaceX AI工程师1小时的课程。

这是我见过最清晰、端到端讲透GrokBot到底该怎么真正用起来的课。

从单agent开始，一路到loop、graph、多智能体协作、verification和agent-friendly架构。

不管你是从没认真跑过agent，还是已经天天在用GrokBot，看完都会把很多“原来如此”的点一次打通。

真正核心就这几层：

1. 99%的人只用1个agent，没用loop和graph，只发挥了1%实力

2. 真正的新stack是Chief of Staff + PM + 20+ workers全自主团队

3. 怎么用高质量skills让agent像真工程师一样干活（pstack）

4. 怎么把代码和架构改成agent-friendly（Dune、contract、CI）

5. 怎么用verification让自己真正敢把活交给它们，而不是一直盯着

大多数人花好几个月才拼出的全景图，这节课直接给你画完。

把时间空出来，这可能是你这个月最值的一堂课。

[Kyle Chan @kyleichan]

Everyone’s so focused on the US-China AI race.

But the fiercer AI race might be Beijing vs. Shanghai vs. Hangzhou vs. Shenzhen.

[Xiaoyin Qu @quxiaoyin]

My bet: @thinkymachines will soon make more money than @AnthropicAI. Not by winning the race to build one standardized frontier model. By becoming the Palantir FDE for enterprise custom models.

The playbook:
1. Release the best American open-weight model.
2. Drive widespread enterprise adoption.
3. Charge the largest companies 7–9 figures to post-train and run custom models behind their own firewall.

The model rests on three bets:
1. Large enterprises will increasingly demand their own models with their own data, and this is how they differentiate and win.
2. Enterprises won’t need just one model. They’ll continuously need new models for different workflows, departments, and proprietary datasets. That creates extremely sticky, recurring revenue.
3. Autoresearch will make custom model development increasingly scalable. Tinker can become the interface enterprises use to post-train their own models—with @thinkymachines providing the expertise and infrastructure behind it. FDE, infra, everything, huge contracts.
4. Eventually, maybe everyone wants their OWN model, and autoresearch and training inside tinker on top of @thinkymachines's base model will make it happen.

Meanwhile, Henry-ford-styled, standardized models will makes no margins. OpenAI and Anthropic will have their API margins squeezed by Deepseek/GLM/Grok/Meta etc, and their consumer subscriptions are loss centers.

The fat margin will move to customization: proprietary data, post-training, evals, deployment, and infrastructure.

If this thesis is right, @thinkymachines isn’t building just another frontier lab. It’s building the highest-value layer between frontier research and enterprise model ownership.

Turns out, the best business model for enterprise is NOT to sell commodity API access. Sell them their own models.

I’m extremely bullish on this approach.
@miramurati may be the most commercially savvy frontier-lab leader. I have to admit it.

[Arc Institute @arcinstitute]

The future of biology is agentic. We're proud to work with NVIDIA on the Evo series of models and are excited to see the NVIDIA BioNeMo Agent Toolkit launch to accelerate programmable biology.

[Yun-Ta Tsai @yunta_tsai]

Many people think any given ML project is 99% training.

In reality, it’s 50% evaluation, 40% data cleaning, 8% integration, and 2% training.

The first two set the noise floor for learning. No ML magic matters; the model cannot lower the noise floor, as that’s the optimal bound of Shannon encoding of your data.

Thus, not a single day goes by without me thinking about ontology. Even the old labels have to be constantly reviewed.

[Shreyas Doshi @shreyas]

https://t.co/BZ9YpF6DOo

[🇮🇹 Massimo De Luisa @massimodeluisa]

Easier way to protect yourself (if you are not infected yet) is to set a minimum release age in your package manager.

For @npmjs:
`npm config set min-release-age=2d`

For @pnpmjs:
`pnpm config set minimumReleaseAge 2880`

For @bunjavascript:
```
# In bunfig.toml
[install]
minimumReleaseAge = 172800
```

For Yarn:
`yarn config set npmMinimalAgeGate "48h"`

[lcamtuf @lcamtuf]

The coreutils Rust rewrite story is pretty funny.

Coreutils are tools like rm, mv, mkdir, etc. Unlike binutils, this isn't a fertile ground for memory safety bugs. But, the rewrite was completed, and in the spirit of progress, Canonical decided to switch.

🡇

[Eric Zhang @ekzhang1]

https://t.co/7zaVlHRBEi is really cool - simple, fast, really easy VMs

Nice that it "just works" - guest kernel is bundled in shared library, and static init process mounts file systems.

The virtiofs server is great too, you don't need a separate virtiofsd. so embeddable! :)

[Boris Cherny @bcherny]

Dogfooding Opus 4.7 the last few weeks, I've been feeling incredibly productive. Sharing a few tips to get more out of 4.7 🧵

[Charlie Marsh @charliermarsh]

We've entered into an agreement to join OpenAI as part of the Codex team.

I'm incredibly proud of the work we've done so far, incredibly grateful to everyone that's supported us, and incredibly excited to keep building tools that make programming feel different.

[Bo Lu @blu3ness]

https://t.co/yBgZmauoHq
it's happening - the machine to machine economy is going to be the next big thing

[Bo Lu @blu3ness]

I'm Building an Agentic Portfolio Advisor https://t.co/1AKCyvIPZG