Expert attention map

The Who's Who of AI

What credible people across AI noticed, why it matters, and where the field is converging or disagreeing.

2,397 searchable experts 4,280 tracked across all sources
Build my stream Create my edition
Filter the conversation Who is saying what?

Combine a professional role with a reaction lens. Both must match the same attributed contribution.

Clear all
New Network reaction maps

See how experts are reacting—not just what they shared.

Posts are grouped by conversation and tone. Select a lens to filter the stream; these are never permanent labels on people.

Opportunity & adoption Research & technical analysis
1 live reaction map in this view Jump to the evidence ↓
Multiple readings Unit distance remarks Research & technical analysis · 2 attributable reactions
Showing signals surfaced by Ben Recht ×
Most important

The developments commanding sustained expert attention

One card per development. Sources are clustered; reaction bundles describe these posts, never the people behind them.

Established AI field signal Signal 4d ago
⚡ 2560 h early

Unit distance remarks

5 experts across 3 network communities independently surfaced this.

5 experts 3 communities 1 sources clustered
Ben Recht Suresh Venkatasubramanian Grace James Grimmelmann Benjamin Riley
8 experts discussed this · 47 posts
Colin: I hate that today’s most prominent AI skeptics keep forcing me to argue against them. I yearn to join them in the fight against the booster zealots, wherein there is still plenty of room for legiti…
Colin: The questions available to a skeptic today are kind of esoteric, like “how much should we as a society be willing to pay for theorems”
Ben Recht: The way forward is through the mushy middle. This technology is amazing. It works and breaks in surprising ways. The main boosters are lunatics. And the loudest critics are detached from reality. F…
Open the full discussion →
Primary source cdn.openai.com ↗
New Safety & security Analysis 5h ago
⚡ 388 h early
alignment artificiality essay

The Artificiality of Alignment

4 experts across 3 network communities independently surfaced this.

4 experts 3 communities 1 sources clustered
Deb Raji Ben Recht Juan Diego Rodriguez Andrea Lathrop

“Anyways, here's some of that discourse from 2019; https://t.co/pzBqIF2MEA + Communities like MD4SG, who piloted in 2016; https://t.co/3CQxeNxYiw And folks like @jessicadai_ explicitly drawing bridges to the language of "alignment" in 2023; https://t.co/3etN…”

“This is very good and was rather prescient: joinreboot.org/p/alignment”

Primary source joinreboot.org ↗
Established AI field signal Analysis 8d ago
⚡ 1900 h early
Reify This ideas letter essay

Reify This

4 experts across 2 network communities independently surfaced this.

4 experts 2 communities 1 sources clustered
Ben Recht Henry Farrell ae Tyler Shoemaker

“Reify This! www.theideasletter.org/essay/reify-...”

“This little sentence in last month’s encyclical helped (=goaded) @leifw.bsky.social, @beenwrekt.bsky.social, and me to tie off some threads that spun out of the Cultural AI conference in March “Reify This”: www.theideasletter.org/essay/reify-...”

3 experts discussed this · 5 posts
Ben Recht: In _The Ideas Letter_, Leif Weatherby, Tyler Shoemaker, and I wrote about the creation of reality through pseudoscience, religion, and psychosis. Of course, the piece is mostly about AI.
Ben Recht: I still can't get over the hero image. Priceless.
Hagen Blix: "If you ask interpretability or alignment researchers how that might come about, you will receive hours-long, mind-numbing lectures about the long-term future of humanity cast in some of the most i…
Open the full discussion →
Primary source theideasletter.org ↗
Established Agents & robotics Research 7d ago
⚡ 3 h early
RL reasoning objective arxiv paper

What is the objective of reasoning with reinforcement learning?

3 experts are actively discussing the implications.

1 expert 1 community 1 sources clustered
Ben Recht

“This is probably too mathy, but the first few sections might help?”

3 experts discussed this · 31 posts
tweety fish: the (low) dimensionality of that thumbs up signal (and the nature of the thumbs up corpus) is the key key key key thing to what LLMs can't do and what they are super weird about and why they're Lik…
tweety fish: this is more about the dimensionality of the RL vector as opposed to the size of the corpus but certainly the fact that generating a high-dimensional and robust RL corpus for RLHF is hellaciously e…
Colin: I’m not, really. I don’t think they’re that different. In RLVR there’s still not a great way to tell the model that the part you liked was when it got the right answer, but you didn’t like the part…
Open the full discussion →
Primary source arxiv.org ↗
Debates gaining attention

What experts are discussing without an anchoring article

6 experts · 47 posts · 5d ago
Ben Recht Luis Villa VE, cybersocial occult investigator Anthony Moser tweety fish Pavel
Reaction converging Questions & unknowns · 3 Building & implementation · 1 Research & technical analysis · 1
tweety fish: like people are getting mad about this but serioulsly, an ethical stand against software developed with AI tools, right now, today, rules out essentially ALL SOFTWARE. I'm sorry! macos. bluesky. th…
tweety fish: anyhow I promise that you have never ever once used a piece of software developed by ME using AI tools, that I can say with 100% confidence.
Open the thread →
4 experts · 29 posts · 3d ago
Ben Recht tweety fish Isaiah Bishop B! 🐝 Cavello (they/them)
Reaction converging Research & technical analysis · 3
Isaiah Bishop: I dont see a reason to think AGI is impossible, but i do think a lot of people are getting over their skis in part because the anti-AI backlash was pretty dumb
Isaiah Bishop: Like my fundamental view is that AI as it exists, is just the kind of embedded capital we have for machines but for knowledge work. Nothing i have seen from them suggests a really extensive interna…
Open the thread →
3 experts · 7 posts · 12d ago
Ben Recht Christopher Mims Amy Hoy
Shared emphasis Research & technical analysis · 2
Christopher Mims: The very first Mac ($2,495 in 1984 dollars) would cost the equivalent of $8,000 today. Cutting-edge prosumer tech is expensive again.
Amy Hoy: The very first Mac ($2,495 in 1984 dollars) would cost the equivalent of $8,000 today. Cutting-edge prosumer tech is expensive again.
Open the thread →