Every example stated is something I am reasonably aware of and several I have written about, barring most things at local/municipal level.
I would assess this is "not a lot of policy impact" in my view, when scaling for 10 years + the size of the community (e.g. number of papers, researchers, time allocated).
So I am not sure what is "revisionist": I think I do account for the things you said. In fact, our claims seem compatible:
"most of the research has failed to actually take the next step and affect policy/public discourse/other forms of consequential decision-making"
and
the negation of "little to no policy impact"
admits a lot of space.
Most research can have ~0 policy impact, a small fraction can have some, and both claims are true.
But the reality is that this is probably the CS conference that still has had the most material impact on what policymakers think & understand about what to do about AI. That's just the historical reality.
Not to mention materially shaping MOST of the internal governance practices at every major AI company today (eg. model cards, harm taxonomies, etc) & still shaping organizational practices, including procurement, in many public interest settings (eg. govt, schools, hospitals).
+ don't even get me started on state/municipal impact!
Did the community adjust to LLMs/"frontier AI" shift perfectly? No. I agree there's a TON of work to do to be adequately responsive to the moment, and I've been disappointed myself by some reactions to the changing landscape
