Science communication of AI/ML research

PR for the robots

2023-12-05 — 2026-04-24

Wherein methods of conveying AI and ML research to public and scientific audiences are surveyed, and a distinction between debate mode and deep-canvassing mode is observed.

academe

communicating

faster pussycat

learning

mind

statistics

Placeholder for ideas about how to communicate AI/ML research to the public and to other scientists.

IMO, it’s deathly important.

I am doing a lot of this right now, especially regarding AI, both safety and economics. I am painfully aware I could improve because I keep stumbling into debate mode rather than deep-canvassing mode.

1 Incoming

Why This OnlyFans Model Posts Machine Learning Explainers to Pornhub; Update: I decided that posting this to the workplace did not satisfy my risk/reward profile
Which AI harms and risks will mobilise the public to act? (Ostarek et al. 2026).

2 References

Gebru, Morgenstern, Vecchione, et al. 2021. “Datasheets for Datasets.” Communications of the ACM.

Mitchell, Wu, Zaldivar, et al. 2019. “Model Cards for Model Reporting.” In Proceedings of the Conference on Fairness, Accountability, and Transparency. FAT* ’19.

Ostarek, Rogers, Kenward, et al. 2026. “Which AI Harms and Risks Will Mobilise the Public to Act?”

Spiegelhalter. 2024. The art of uncertainty: how to navigate chance, ignorance, risk and luck ; a Pelican book. Pelican books.