AI is breaking our proxies for expertise

kept by novawanderer sep 13

AI’s success in solving high‑profile math problems threatens the puzzle‑solving proxy that signals mathematical expertise, risking erosion of genuine idea generation and progress.

I Had Claude Read Every AI Safety Paper Since 2020, Here's the DB — LessWrong
Claude AI read ~4000 AI safety papers since 2020 and created a searchable database with summaries, tags, and metadata to help researchers find relevant work.
also on arXiv, #open-source, #developer-tools
SEO vs AEO vs GEO vs AIO: Same Work, Different Acronyms
AEO, GEO, AIO, LLMO, and AI SEO describe the same optimization work; real distinctions are fetch‑vs‑training knowledge and mention‑vs‑citation, not the acronyms themselves.
also on arXiv, #ai, #developer-tools
AI is doing something weird to Science
Scientific discovery emerges from a loop of human question-posing, LLM proposal, external verification, and human curation, not from AI alone.
also on Terence Tao, Lean, #open-source
LLMs reward expertise
The article argues that domain expertise, not just prompting skill, determines how much value users can extract from LLMs, as experts steer models more effectively.
also on Terence Tao, OpenAI, Anthropic, #ai-coding, #developer-tools
Human mathematicians are being outcounterexampled
AI tools have rapidly generated and formalized counterexamples to longstanding conjectures such as Erdős’ unit distance, Grothendieck’s group scheme problem, and the Jacobian conjecture, showing machines now out‑counterexample humans.
also on Lean, OpenAI, #open-source, #ai-coding, #developer-tools