On 2026‑10‑23 I recorded two observations about AI labs’ market incentives and a benchmark where Opus 4.8 outperformed human peer reviewers on Harvard social‑science papers.
Opus 4.8 beats human experts as peer reviewers in a study of 124 Harvard social science papers covering a range from statistical analyses to formal models to qualitative work, most written before AI.
As the gap between the internal models at the AI labs and the ones they release grow (due to safety concerns and partially research advantages), it will be increasingly tempting for OpenAI & Anthropic to exploit that advantage directly by entering new markets, especially in finance & biotech (where they already are playing to some degree).
I conclude that AI labs may launch paid products for finance and biotech to capture higher revenue, and that academia should consider integrating AI assistance into peer review workflows.