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.

Research & technical analysis
2 live reaction maps in this view Jump to the evidence ↓
Reaction converging How LLMs Distort Our Written Language Research & technical analysis · 3 attributable reactions Shared emphasis Training AI Scientists to Replicate Research Research & technical analysis · 2 attributable reactions
Showing signals surfaced by Eugene Vinitsky ×
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.

Developing Models & releases Research 2d ago
⚡ 4 h early
LLMs distort written language paper

How LLMs Distort Our Written Language

4 experts across 2 network communities independently surfaced this.

4 experts 2 communities 1 sources clustered
Eugene Vinitsky Ben Williamson Isaiah Bishop Jed Brown
How the network is reacting 3 experts are independently emphasizing research & technical analysis.
Reaction converging

Research & technical analysis

3 experts

Evidence, methods and technical implications.

“Fantastic paper demonstrating how LLM editing is gradually distorting our writing and the way we think to write: arxiv.org/abs/2603.18161”

3 experts discussed this · 4 posts
Eugene Vinitsky: Fantastic paper demonstrating how LLM editing is gradually distorting our writing and the way we think to write: arxiv.org/abs/2603.18161
Eugene Vinitsky: Sitting in Marwa's fantastic PhD defense, trying my best not to just retweet all her papers
Isaiah Bishop: Fantastic paper demonstrating how LLM editing is gradually distorting our writing and the way we think to write: arxiv.org/abs/2603.18161
Open the full discussion →
Primary source arxiv.org ↗
Established AI field signal Signal 19d ago
⚡ 274 h early

A Compositional Framework for Open-ended Intelligence

2 experts are actively discussing the implications.

2 experts 2 communities 1 sources clustered
Eugene Vinitsky Ida Momennejad

“In the neuro->AI direction cog sci insights can help shift 1-AI evaluation from benchmarks to adaptive paradigms 2-AI architecture from input-output/next-token objectives to pre-/post-training over algorithmic operations/ grammar. We're developing next-prim…”

2 experts discussed this · 7 posts
Ida Momennejad: New work with Roberta Raileanu: A Compositional Framework for Open-ended Intelligence Open-ended intelligence is the capacity to adapt to novel problems & environments that are substantially differ…
Ida Momennejad: A mathematics of open-ended intelligence needs 2 pillars: 1. a minimal set of representational primitives and algorithmic primitives 2. an acquired compositional grammar for selection, recursion, &…
Ida Momennejad: We formalize open-ended intelligence in terms of the compositional closure induced by a finite primitive set P & a set of composition operators C. We characterize properties of the induced closure …
Open the full discussion →
Primary source arxiv.org ↗
Established AI research Research 16d ago
⚡ 198 h early
AI scientist research replication training

Training AI Scientists to Replicate Research

2 directory members surfaced this signal.

2 experts 2 communities 1 sources clustered
Eugene Vinitsky Sung Kim
How the network is reacting 2 experts are emphasizing research & technical analysis.
Shared emphasis

Research & technical analysis

2 experts

Evidence, methods and technical implications.

“AI for science research paper with a limited time and compute budget, and without access to the original plot. Blog: inherentlabs.ai/research/tra... Paper: arxiv.org/abs/2608.13331 Model:”

Primary source arxiv.org ↗
Established Agents & robotics Analysis 9d ago
⚡ 40 h early
RL evaluation bottleneck analysis

RL is bottlenecked by evaluation, not algorithms - rl-blogging

2 experts are actively discussing the implications.

2 experts 2 communities 1 sources clustered
Eugene Vinitsky Miguel Alonso Jr.

“If like me you're curious why most algorithmic papers in RL land with a thud, here's my argument as to why.”

2 experts discussed this · 4 posts
Eugene Vinitsky: If like me you're curious why most algorithmic papers in RL land with a thud, here's my argument as to why.
Eugene Vinitsky: As always my rule is write fast with minimal editing, so forgive the mistakes. They at least prove minimal LLM involvement.
Miguel Alonso Jr.: If like me you're curious why most algorithmic papers in RL land with a thud, here's my argument as to why.
Open the full discussion →
Primary source rl-blogging.leaflet.pub ↗
Established Models & releases Signal 10d ago
⚡ 3 h early

LLMs Can Design Near-Optimal OR Algorithms

2 experts are actively discussing the implications.

1 expert 1 community 1 sources clustered
Eugene Vinitsky
2 experts discussed this · 3 posts
Eugene Vinitsky: Well arxiv.org/abs/2608.27296
angela zhou: very nice and interesting study! To be fair though the algorithms to the benchmarks are also well-known in the literature. maybe some are paywalled
Eugene Vinitsky: Yeah, I think the main takeaway is more that if you're able to formalize the problem, which of course is already somewhat a task for experts, you can get quite far away after that falling back on L…
Open the full discussion →
Primary source arxiv.org ↗
Established Evaluation & benchmarks Research 13d ago
⚡ 5 h early
reward hacking terminal task benchmark

Hack-Verifiable Terminal Bench: Evaluating Reward Hacking in Terminal Tasks

1 directory member surfaced this signal.

1 expert 1 community 1 sources clustered
Eugene Vinitsky

“You want to check if your agents are going to reward hack but how do you know when that happens? This paper builds instantly verifiable instances of reward hacking through honeypots in terminal-bench: arxiv.org/abs/2608.22103”

Primary source arxiv.org ↗
Established AI field signal Analysis 20d ago
⚡ 17 h early
social media algorithm transparency essay

Quit social media that won't tell you their algorithm

2 experts are actively discussing the implications.

2 experts 1 community 1 sources clustered
Eugene Vinitsky zzstoatzz.io

“Still think being able to understand the algorithm you're interacting with is very underrated as important www.eugenevinitsky.com/posts/audita...”

2 experts discussed this · 7 posts
Eugene Vinitsky: Still think being able to understand the algorithm you're interacting with is very underrated as important www.eugenevinitsky.com/posts/audita...
Eugene Vinitsky: also love it when I read something I wrote months later and think it holds up. Good feeling.
Eugene Vinitsky: It's good!
Open the full discussion →
Primary source eugenevinitsky.com ↗
Debates gaining attention

What experts are discussing without an anchoring article

7 experts · 28 posts · 1d ago
Eugene Vinitsky Daniel Lowd Eli Sennesh James Tauber rev. howard arson Evangelos Kazakos brennan
Multiple readings Opportunity & adoption · 2 Building & implementation · 1 Questions & unknowns · 1
Eugene Vinitsky: (1) Artists are justified in being angry about AI and we really shouldn’t use AI art where otherwise we would have hired an artist. Art isn’t really about productivity growth and automating it help…
Evangelos Kazakos: (1) Artists are justified in being angry about AI and we really shouldn’t use AI art where otherwise we would have hired an artist. Art isn’t really about productivity growth and automating it help…
Open the thread →
6 experts · 21 posts · 9d ago
Mark Riedl Eugene Vinitsky Juan Diego Rodriguez Colin rev. howard arson Sophie Greenwood
Multiple readings Research & technical analysis · 2 Building & implementation · 1 Questions & unknowns · 1
Eugene Vinitsky: Repulsed by the idea that we should sit idly by while frontier labs determine the trajectory of the AI transition. They don’t know how to make systems “safe” or “aligned“ nor do any such notions ma…
Eugene Vinitsky: I think, without ever acknowledging it explicitly, I had sort of resigned myself to this. But now I feel weirdly re-inspired to do some useful work here
Open the thread →
5 experts · 19 posts · 17d ago
Eugene Vinitsky Luis Villa Colin Isaiah Bishop Andrea Lathrop
Multiple readings Questions & unknowns · 2 Concern & critique · 1 Research & technical analysis · 1
Eugene Vinitsky: The main reason AI companies went with the obviously bad PR strategy of “AI will take all jobs“ is that they believe it and it worries them
Eugene Vinitsky: This is a much more parsimonious hypothesis than “these people are too stupid to understand that this campaign will incur backlash”
Open the thread →