Google's Co-Scientist Upgrade: AI Controls Real Lab Equipment to Synthesize Materials
新智元 · wechat · 2026-08-31
Google DeepMind and Duke University released a major upgrade to the AI research system Co-Scientist, achieving a closed loop from hypothesis generation to real-world experimental execution. In materials science, AI-designed recipes successfully synthesized novel 2D carbide crystals and single-layer semiconductors (e.g., MoS₂) using lab-specific equipment, exploring non-toxic routes. Validations were also achieved in biology and CS. The system incorporates anti-hallucination and anti-plagiarism mechanisms. This marks a pivotal shift for AI research from computational thinking to physical verification.
Related event: Google's Co-Scientist Achieves Closed-Loop Discovery Across Sciences(2 posts)→
More from Research
- A beginner-friendly introduction to causal inference kicks off, Part 1 — Nasereliver · 2026-10-16
- FBI method trains for Pass@K with one sample per update at 12.3% of the cost — EmtiyazKhan · 2026-10-16
- Berkeley researchers sample humanoid robot policies directly from weight space with diffusion — berkeley_ai · 2026-10-16
- OneWM: unified causal training shows world and action modeling improve each other — IAAI233 · 2026-10-16
- Baidu's Virtual Head Attention tames the KV cache, FLOPs, quality triangle in long-context LLMs — baidu · 2026-10-16
- 5.8M-param Speech Event Detector adds laughs, breaths and背景 sounds to Whisper transcripts — Ayush Kumar Bar · 2026-10-16