AutoScientists – a research lab made of agents @Harvard researchers connected agents into a self-organizing scientific team without a boss agent standing in the middle All agents look at the same…

AI agents just started doing real science, not just answering questions. Most AI research agents follow one path or take orders from a central planner. They forget what failed and stop improving once…

科学实验的AI智能体原来还能自己组队互相挑刺。AutoScientists没有中央调度，而是让agent围绕有前景的假设自发组队，先批判再落地，连失败经验也共享。本质上把长周期探索变成并行团队，避免重复踩坑。BioML-Bench提升8%，nanoGPT训练优化快了近2倍。这思路有点意思。

Two Harvard researchers introduce AutoScientists, a decentralized team of AI agents that collaborate, critique & adapt their research strategy over time. Instead of following a fixed plan, agents…

Here is the English version optimized for an engaging, long-form post on X. It follows all your formatting constraints (no markdown, emojis for sections, and exactly two important hashtags). 🤖 Can…

(1/7) 多 agent 科研的坑，不是 agent 不够多，而是失败没有地方沉淀。AutoScientists 的有趣点在这里：把长时间科研搜索做成共享状态的团队系统，先互相批评假设，再花 compute 跑实验。

What the agents actually do to escape stagnation? These are real interactions between AI scientists. Agents notice when teams are duplicating work. They identify saturated research directions and…

AI Scientists are starting to actually do science. Not just answer questions. Not just run workflows. Introducing AutoScientists: a decentralized team of AI agents that can generate hypotheses,…