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
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.

Showing signals surfaced by David Marx ×

The developments commanding sustained expert attention

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

Established Agents & robotics Signal 16d ago
⚡ 1924 h early

GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

3 experts across 3 network communities independently surfaced this.

3 experts 3 communities 1 sources clustered
Established AI research Research 4d ago
⚡ 195 h early
AI and consciousness paper

AI and Consciousness

3 experts across 3 network communities independently surfaced this.

3 experts 3 communities 1 sources clustered

““Genuine conscious experience might require something no AI system could possess for the foreseeable future – intricate biological processes, for example, that silicon chips could never replicate.” www.cambridge.org/core/element...”

“free download for 2 weeks by @eschwitz.bsky.social”

Established AI research Research 3d ago
⚡ 5 h early
language-informed loss visual generation paper

A Language-Informed Loss for Visual Generation

2 directory members surfaced this signal.

2 experts 1 community 1 sources clustered

“Researchers have unveiled LILIE, a Language-Informed and Lossless Encoder that boosts text-to-image generation with language supervision, greatly reducing errors. This method no longer needs lossy tokenizers, providing a stronger training signal than pixel …”

Established AI field signal Signal 10d ago

The Limiting Dynamics of SGD: Modified Loss, Phase Space Oscillations, and Anomalous Diffusion

2 experts are actively discussing the implications.

1 expert 1 community 1 sources clustered

“now do a version where instead of converging to a point minimum, they orbit a line minimum while continuing to move further from the origin.”

2 experts discussed this · 3 posts
David Marx: www.youtube.com/watch?v=Vv3f...
David Picard: A little Hollywood treatment for my #DeepLearning course 🎬 Coming Winter 2027.
David Marx: now do a version where instead of converging to a point minimum, they orbit a line minimum while continuing to move further from the origin.
Open the full discussion →
Established AI field signal Signal 16d ago

Ai2 Asta

2 experts are actively discussing the implications.

1 expert 1 community 1 sources clustered

“oh boy have I got a toy for you asta.allen.ai/discover”

2 experts discussed this · 23 posts
Joshua Grochow: Can one do "algebraic geometry" over a monoid in place of the coordinate ring? That is, suppose I have a commutative monoid. Is there a coherent (in the informal sense) notion of Zariski topology? …
Joshua Grochow: (Answers would be great, but I'm more looking even for keywords, authors, etc to look up - hard to search for.)
David Marx: oh boy have I got a toy for you asta.allen.ai/discover
Open the full discussion →

What experts are discussing without an anchoring article

3 experts · 7 posts · 1d ago
Multiple readings Research & technical analysis · 1 Questions & unknowns · 1
Ted Underwood: The Internet era is kind of rough on faculty in emerging fields, because if University U hasn’t yet hired someone in your field, the students there *will write you* to ask sincere, interesting ques…
Ted Underwood: I guess my takeaway is not, students pls be less curious & proactive — but more hurry the f up, Univ U.
Open the thread →
3 experts · 10 posts · 20d ago
Multiple readings Research & technical analysis · 1 Opportunity & adoption · 1 Questions & unknowns · 1
Berna Devezer: i cannot wrap my head around that 2 papers a week (socarxiv) or a month (arxiv) should constitute a meaningful cap on preprint submissions. 2 papers a year (no one) though? that i'd get.
David Marx: They probably picked that threshold based on the current percentiles to try to at least "desk reject" the worst offenders. powerlaw effect as usual: it's likely a minority of bad actors are committ…
Open the thread →