ICWM Adopts Single-Blind Review to Publicly Expose AI Slop Authors

ylecun · x · 2026-09-23

Yann LeCun retweeted Randall Balestrieri about the International Conference on World Modeling (ICWM): the conference uses single-blind review and is entirely open to anyone (like ICLR in its early days), so that authors producing AI slop are publicly identified, while anonymous reviewers stay free of peer pressure.

In the quoted thread, Phillip Isola pushes back: ICLR is double-blind even for senior roles — is this referring to other conferences? He agrees slop should be penalized but is unsure removing blind review is the best option.

Related event: LeCun Backs New ICWM Conference Targeting Bloated AI Venues(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →