How AI is affecting the practice of science.
Latest August 17, 2026 No. 021 AI peer review ensembles, Riemann zeta bound, VibeMathed registry The verification thread keeps tightening. Anthropic's Claude improved a Riemann zeta bound and formalized the proof in Lean. VibeMathed now catalogs 565 AI-solved math problems, 134 of them Lean-verif...Get it in your inbox
Archive
- Astra solves ten open math problems, AI agents fail at open-ended research, peer review under strain
- AlphaFold team disbanded, verification as the bottleneck, Tao's ICM caution
- Genesis Mission $5B, Challengescape problem directory, team science vs. solo AI agents
- Science EIC contra AI hype, data poisoning of research agents, Vancouver Standard for AI disclosure
- Biomni in Science, Co-Scientist in Nature, open-vs-closed AI workbenches
- Claude Science workbench, Google's Paper Assistant Tool, NeurIPS hidden review prompts
- Science's AI productivity paradox, Sakana clears peer review, AI proof search on open Erdős problems
- Provenance vocabulary, Wikipedia contamination, OpenAI rare-disease diagnoses, ARA paper format
- Claude Fable/Mythos, Leiden Declaration, Frontiers editor resignation
- Sum-product conjecture disproof, AI-generated authorship fraud, OMB grant-cost rules
- AutoScientists governance, AI-agent monoculture, nonsense-term retractions
- AI agents p-hacking, submission surge data, Gemini for Science
- AI co-mathematician, compiler-verified protocols, open infrastructure bets
- Fabricated citations at scale, vibe physics, zero-click readership
- AAAI-26 AI review pilot, ScienceOne 100, NEJM retraction
- GPT-Rosalind, AI verification gap, Agent4Science
- HAI AI Index, virtual cell automation, MCP for scientific literature
- Bixonimania hoax, Google's paper-writing agents, AlphaFold expands the frontier
- Hallucinated citations at scale, ICML catches AI reviewers, scholarly infrastructure gaps
- arXiv goes independent, AI Scientist v2, Claude Scholar
- What this is, why it exists, who writes it