The geopolitical angle of the AI race is only going to keep accelerating. More domestic talent travel restrictions from the Chinese government. I expect more and more things like this, AI is getting embedded in the core of existing power structures. www.bloomberg.com/news/arti…
Who's Who of AI
Nathan Lambert
Post-training researcher at Ai2, writes Interconnects
A LLN - large language Nathan - (RL, RLHF, society, robotics), athlete, yogi, chef
Writes http://interconnects.ai
At Ai2 via HuggingFace, Berkeley, and normal places
What they're sharing
China Expands Travel Curbs to Top AI Talent at Private Firms bloomberg.com
interconnects.ai
interconnects.ai
interconnects.ai
interconnects.ai
interconnects.ai
interconnects.ai
interconnects.ai
interconnects.ai
chinai.substack.com
Articles & links
Why I think Anthropic's uneven safety policies with the release of Claude Fable 5 undermine the broader AI community's cohesion and accelerate us to more uncertainty and risk in AI's near-term evolution. www.interconnects.ai/p/claude-fab...
Some ideas for what comes next, May 2026 Gemini Flash 3.5, Mythos, open-closed balance, America's open-source surge, emerging power struggles and more. www.interconnects.ai/p/some-ideas...
My time at Ai2 / @ai2.bsky.social has come to an end. Ai2 is a wonderful place. The last 2.5+ years building Olmo, Tulu, and other projects will be one of the peaks of my entire career. www.interconnects.ai/p/farewell-ai2
Welcome to the AGI era of AI governance It's a one-way door and we weren't ready for it. Especially the open-source community celebrating Anthropic's downfall isn't ready for when it's their turn in court. www.interconnects.ai/p/welcome-to...
Latest open artifacts (#21): Open model bonanza! Gemma 4, DeepSeek V4, Kimi K2.6, MiMo 2.5, GLM-5.1 & others. On CAISI's V4 assessment. An eventful month with one flagship release after another www.interconnects.ai/p/latest-ope...
AI Weekly's analysis
→
- DeepSeek-V4-Flash ranked best-in-class for local agentic coding tasks across the full May open-model cohort benchmarked by Lambert.
- Poolside's Laguna XS.2, a 33B MoE model under Apache 2.0, is the strongest open-weight release at its size class in this cohort.
- The open-to-closed capability gap stands at 3-7 months and has been narrowing since DeepSeek R1 launched, per CASI data.
Read full analysis →
Can closed labs can create margins with very high cost services in the long-term? I say yes, but only in a few niches like high end knowledge work. Meanwhile, the open model economy slowly grows to be bigger and control the long-tail of AI use. www.interconnects.ai/p/open-and-…
I wrote a little state of the union on the state of Interconnects, how I ended up with a loyal niche audience, and how I’m tweaking operations to ensure the long-term success here! Thanks for the support everyone. 💚
Interconnects: www.interconnects.ai/p/frontier-p...
Very useful post from @jjding99 on Anthropic's China Dogma chinai.substack.com/p/anthropics...
Their own posts
Recent commentary
Being out of SF has lowered my information proximity but with the big upside of giving me space to cultivate my own beliefs and values around ai. We need more people zagging in AI, the monoculture just helps the incumbents win at this point.
It's been a great effort by the early and growing American open-model labs since last June to put the US much more back on the map. We were getting totally owned last June. Nvidia, Ai2, Arcee, Gemma, GPT-OSS and a few others will be seen as saving American open AI.
On-policy distillation is on track to be a lasting method in post-training. The list of areas would be: Instruction tuning (SFT/IFT) RLHF Direct Preference Optimization (DPO et al) RLVR On-policy Distillation (OPD) New classes of methods are rare! Excited to play.
Their network
In Nathan Lambert's orbit
Center = Nathan Lambert. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.