AI researchers question conference peer review as blogs and repos become the real publishing channel

giuliacfanti · x · 2026-11-16

Dimitris Papail argues much of the best recent AI research now appears as blog posts, GitHub repos and X threads rather than conference papers, calling the review stamp requirement meaningless. Victor Veitch says he already advises students to be ambitious, post on GitHub and Twitter, and minimize friction to doing and sharing excellent work. Giulia Fanti raises the practical concern: can students still graduate and get hired for research positions without conference publications? The thread captures growing distrust of the traditional peer-review pipeline in AI research.

Related event: DeepMind researcher declares the "publish for conferences" game dead(9 posts)→

Original post →

More from AGI Musings

AGI Musings channel →