31. What AI Safety needs is better tooling and engineering. For instance, building stronger red-teaming... What AI Safety needs is better tooling and engineering. For instance, building stronger red-teaming tools. Great to see this open-source effort to red-team image models. It seems t 10/15 01:45 ♡
32. Another great application of System One models. I think much of what we call hallucination in web ... Another great application of System One models. I think much of what we call hallucination in web agents comes from misreading the page. When a page is flattened to text, a small ✗ 10/15 00:16 ♡
33. Agents will do more and more of their work inside systems of record, so those systems need to be bui... Agents will do more and more of their work inside systems of record, so those systems need to be built for them. Attio's new hybrid architecture is one of the cleaner approache 10/14 23:55 ♡
34. Multiplayer AI is what powers successful businesses today. It enables self-improving businesses in... Multiplayer AI is what powers successful businesses today. It enables self-improving businesses in which human and personalized agents collaborate. @zenzap_chat puts people and AI 10/14 23:46 ♡
35. Own your intelligence stack, folks! If you train or fine-tune your own models, you also need a way ... Own your intelligence stack, folks! If you train or fine-tune your own models, you also need a way to prove they are yours once the weights are out in the world. Model fingerprinti 10/14 22:40 ♡
36. Interesting new research from Meta Superintelligence Labs It discusses how reward hacking affects c... Interesting new research from Meta Superintelligence Labs It discusses how reward hacking affects coordination in multi-agent coding systems. Coding agents on a team share code, me 10/14 20:30 ♡
37. Learn to build custom harnesses, folks! Here is an impressive paper showing the impact of a good cu... Learn to build custom harnesses, folks! Here is an impressive paper showing the impact of a good custom harness on long-context work. GLM-5.3-Flash goes from 4.4% to 61.1% exact ma 10/14 03:00 ♡
38. Go-to-market is quickly becoming an engineering problem. Good outreach depends on clean lead data a... Go-to-market is quickly becoming an engineering problem. Good outreach depends on clean lead data and on giving your AI agent the right context about who buys and why. Pancake buil 10/14 01:39 ♡
39. Pay close attention to harnesses, part 2. Last week I shared a paper where a model more than double... Pay close attention to harnesses, part 2. Last week I shared a paper where a model more than doubled its score after training inside the harness it runs in. The same idea applies a 10/14 00:57 ♡
40. B Browse the corpora at https://t.co/KLSajbBNKV Browse the corpora at hub.xerj.org 💬 0 🔄 1 ❤️ 3 👀 1378 📊 1 ⚡ Powered by xgo.ing 10/14 00:50 ♡
