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.

2,364 searchable experts 2,966 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
Active evidence filter

Showing developments with attributable Concern & critique reactions.

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.

Concern & critique Opportunity & adoption
1 live reaction map in this view Jump to the evidence ↓
Multiple readings Security incident disclosure — July 2026 Concern & critique · 2 attributable reactions
Showing signals surfaced by Matthew Sheffield ×
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 Safety & security Development 7d ago
HuggingFace security incident July 2026

Security incident disclosure — July 2026

5 experts across 4 network communities independently surfaced this.

Why this matches Concern & critique reaction 1 attributable expert contribution · Leon Derczynski
“"When we started analysis, we used commercial APIs. This did not work: requests were blocked by the providers' safety guardrails. We ran the forensic analysis instead on our own infra: no attacker data, and none of the credentials it referenced, left our en…” evidence ↗
5 experts 4 communities 1 sources clustered
Maria Antoniak Liz Fong-Jones (方禮真) Stephen Turner Leon Derczynski Matthew Sheffield
How the network is reacting Experts are approaching this through 2 distinct lenses.
Multiple readings

Concern & critique

1 expert

Risks, limits and unintended consequences.

“"When we started analysis, we used commercial APIs. This did not work: requests were blocked by the providers' safety guardrails. We ran the forensic analysis instead on our own infra: no attacker data, and none of the credentials it referenced, left our en…”

Opportunity & adoption

1 expert

New capabilities, benefits and practical upside.

“These programs are made from humanity's collective efforts. It's vital that their code and weights be available to all human beings, not just so that progress can be iterated faster, but also so that bad actors can be stopped faster as well. Great example h…”

Primary source huggingface.co ↗
Established Models & releases Research 15d ago
⚡ 871 h early
LLMs introspection reality check paper

Can LLMs Introspect? A Reality Check

2 directory members surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · Matthew Sheffield
“Some AI interpretability researchers seem prone to mistake externally attributed function for internal methods. In their haste to pronounce that LLMs can "introspect," some researchers fail to rule out more simple descriptions. Great paper on this subject:” evidence ↗
2 experts 2 communities 1 sources clustered
Shauli Ravfogel Matthew Sheffield

“Some AI interpretability researchers seem prone to mistake externally attributed function for internal methods. In their haste to pronounce that LLMs can "introspect," some researchers fail to rule out more simple descriptions. Great paper on this subject:”

Primary source arxiv.org ↗
Established Policy & governance Development 13d ago
Chinese users farewell AI companion apps

'Like my lover': Chinese users bid farewell to AI companions

1 directory member surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · Matthew Sheffield
“China has banned companies from making AI companion services out of concern that they damage humans' abilities to relate to each other www.yahoo.com/news/world/a...” evidence ↗
1 expert 1 community 1 sources clustered
Matthew Sheffield

“China has banned companies from making AI companion services out of concern that they damage humans' abilities to relate to each other www.yahoo.com/news/world/a...”

Primary source yahoo.com ↗
Debates gaining attention

What experts are discussing without an anchoring article

4 experts · 46 posts · 7d ago
Ted Underwood Pwnallthethings SE Gyges Matthew Sheffield
Matches Concern & critique
“Exploiting software is a classic dual-use capability. Exploiting software can be useful to break into systems, but it can also be useful for finding weaknesses in software to fix those issues. Beca…” evidence ↗
Multiple readings Research & technical analysis · 2 Markets & investment · 1
Pwnallthethings: This is going to be a big deal, but easy to get the wrong end of the stick on it. So I think it's worth breaking down what actually happened, why, and what it actually means, based on the public in…
Pwnallthethings: Upfront tl;dr: a model at OAI hacked out of a constrained environment inside OAI and hacked a *different* company autonomously, without authorization from any human in order to creatively solve a t…
Open the thread →