BrainGPT creator: AI discoveries will soon be as explainable as quantum mechanics to a dog
every · x · 2026-09-15
Connor Love (@ProfData), who helped build BrainGPT — an LLM that predicts neuroscience results before experiments are run — argues AI will eventually produce scientific findings too complex for humans to fully explain, "like explaining quantum mechanics to a dog." That will force new explanation methods, such as training models on two related fields to see if one predicts the other and treating that correlation as the explanation. Full discussion in the AI & I episode.
More from AGI Musings
- Andrew Dai: Next-token prediction scales for language, but video needs a new objective — AndrewDai · 2026-09-15
- Alison Gopnik highlights piece linking LLMs as cultural technology to AI math progress — AlisonGopnik · 2026-09-15
- "Not knowing if we're at the start of millions of years or the last decades" goes viral — birchlse · 2026-09-15
- AI CEOs Preach Slowing Down While Anthropic, SoftBank, xAI Race Ahead — eyishazyer · 2026-09-15
- New paper argues RSI isn't near: agents fail to reproduce unpublished NeurIPS work — we_are_mammals · 2026-09-15
- Mathematician littmath's Essay Argues Math Should Shift From Producing to Understanding in the AI Era — littmath · 2026-09-15