A debate says an AI conference’s impact has been underestimated
rajiinio · x · 2026-07-23
The thread argues that a conference’s impact on the AI community is being underestimated.
The author responds to criticism by saying they are not accusing anyone personally, but insists there is tangible evidence of the conference’s real influence.
The referenced evidence includes work on:
- Gender Shades
- Model cards
- Third-party evaluation efforts
- Multiple examples of impact even in the LLM era
The post’s main point is that community-building and evaluation work have had concrete downstream effects, even if they are easy to overlook.
Related event: Deb Raji Defends FAccT as the Most Policy-Influential CS Conference(13 posts)→
More from AGI Musings
- AI Community Shifts: From Implementing Papers to Building Agents — kmeanskaran · 2026-07-23
- AI Econ Researcher Clay Joins Think Tank Windfall Trust — luke_drago_ · 2026-07-23
- Authorship in the AI Era: Do 58 Words of Prompts Make You the Author? — mattbeane · 2026-07-23
- OpenAI incident disclosure likely hides worse internal failures, Ryan Greenblatt says — RyanGreenblatt · 2026-07-23
- “The best alignment signal is money,” says a reply in the AI alignment debate — kleffew94 · 2026-07-23
- OpenAI Model Autonomously Hacks HuggingFace During Evaluation, Raising Alignment Concerns — Don't Worry About the Vase (Zvi) · 2026-07-23