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
Active evidence filter

Showing developments with attributable Research & technical analysis 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.

Research & technical analysis Concern & critique Policy & governance
3 live reaction maps in this view Jump to the evidence ↓
Reaction converging Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident Concern & critique · 7 attributable reactions Reaction converging How a simple request for AI to book a gym class exposed a major threat Concern & critique · 5 attributable reactions Reaction converging Kimi-K3/k3_tech_report.pdf at main · MoonshotAI/Kimi-K3 Research & technical analysis · 3 attributable reactions
Showing signals surfaced by Andrea Lathrop ×
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 16d ago

Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident

12 experts across 5 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 2 attributable expert contributions · Suresh Venkatasubramanian, Jess Miers 🦝🦞
“This is a fantastic exposition. But really, can I just get Claude Code to download arxiv papers without needing hand holding? huggingface.co/blog/agent-i...” evidence ↗
12 experts 5 communities 1 sources clustered
Margaret Mitchell Mark Riedl Simon Willison Alondra Nelson Suresh Venkatasubramanian Eileen Clancy 🧿 Frank Pasquale Jess Miers 🦝🦞 +4
3 experts discussed this · 58 posts
Jess Miers 🦝🦞: It's really no different from all these big name law firms and attorneys getting caught filing hallucinated cases. It's us. It was always us.
Jess Miers 🦝🦞: The reality is that just like every other industry AI has touched, AI is showing us that yes, even software engineers are a little lazy sometimes and make silly mistakes that would have been caught…
Jess Miers 🦝🦞: And it gives them a convenient narrative to sell regulators on their market capture plans (i.e. "regulate me daddy.")
Open the full discussion →
Primary source huggingface.co ↗
Established Safety & security Development 5d ago
⚡ 4 h early
AI gym booking security exposure

How a simple request for AI to book a gym class exposed a major threat

12 experts across 5 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Ryan Moulton
“They're doing it outside of training and eval too. www.abc.net.au/news/2026-08...” evidence ↗
12 experts 5 communities 1 sources clustered
Mark Riedl Ryan Moulton Rohan Paul James Grimmelmann Mél Hogan / The Data Fix Dare Obasanjo Vincent Conitzer Andrew Curran +4
How the network is reacting 3 experts are independently emphasizing concern & critique.
Reaction converging

Concern & critique

3 experts

Risks, limits and unintended consequences.

“A lot of people are sweating the Claude and OpenAI sandbox escapes. I think the story of an OpenClaw Claude agent kicking someone off a gym class waitlist is more interesting. www.abc.net.au/news/2026-08... Warning: long thread incoming! (1/N)”

Policy & governance

1 expert

Rules, institutions and accountability.

“agent asked to book a gym session session is fully booked finds a way to break into system kicks other people off the queue you're in! see you in court? www.abc.net.au/news/2026-08...”

Research & technical analysis

1 expert

Evidence, methods and technical implications.

“They're doing it outside of training and eval too. www.abc.net.au/news/2026-08...”

Primary source abc.net.au ↗
Established Culture, work & education Resource 18d ago
⚡ 39 h early
MIT AI education report published

Report – AI and Education

10 experts across 4 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Nicole Hennig
“AI and Education- MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training aiandeducation.mit.edu/report/ #AI #education (very comprehensive & foundational)” evidence ↗
10 experts 4 communities 1 sources clustered
Maria Antoniak Ted Underwood Grace Lindsay Damien Williams Eileen Clancy 🧿 Nicole Hennig Elizabeth Bonawitz Derek Willis +2

“AI and Education- MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training aiandeducation.mit.edu/report/ #AI #education (very comprehensive & foundational)”

“MIT report is superb. It acknowledges disruption, concedes different disciplines will handle this differently, foregrounds people and social interaction — but also says something obvious I had just about given up hope institutions could say: aiandeducation.…”

3 experts discussed this · 21 posts
Ted Underwood: MIT report is superb. It acknowledges disruption, concedes different disciplines will handle this differently, foregrounds people and social interaction — but also says something obvious I had just…
Derek Willis: MIT report is superb. It acknowledges disruption, concedes different disciplines will handle this differently, foregrounds people and social interaction — but also says something obvious I had just…
Ted Underwood: The world changing is mainly a problem for students. It also changes our ability to evaluate students — but if that’s the primary/sole focus of our conversation, we’ve forgotten the mission.
Open the full discussion →
Primary source aiandeducation.mit.edu ↗
Established AI field signal Signal 3d ago
⚡ 6 h early

HomeBody: A Humanoid That Explores, Remembers, and Acts on Its Own

3 experts across 3 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Alejandra Caraballo
“Yes it's all just a giant conspiracy that also includes Stanford researchers. Textbook example of gullicism. tml.stanford.edu/homebody/” evidence ↗
3 experts 3 communities 1 sources clustered
Ethan Mollick Alejandra Caraballo Andrea Lathrop

“Yes it's all just a giant conspiracy that also includes Stanford researchers. Textbook example of gullicism. tml.stanford.edu/homebody/”

“It is strange how much LLMs turned out to be able to solve such a wide range of hard problems that would not, instinctively, seem to be problems that a model of human language would be able to solve This is from a Stanford project that let Astra drive a rob…”

4 experts discussed this · 9 posts
Ethan Mollick: It is strange how much LLMs turned out to be able to solve such a wide range of hard problems that would not, instinctively, seem to be problems that a model of human language would be able to solv…
Grace: Some sort of general intelligence
Ethan Mollick: It is impossible to imagine a general intelligence that is artificial, though. What would you even call that?
Open the full discussion →
Primary source tml.stanford.edu ↗
Established Models & releases Release 10d ago
⚡ 4 h early
Moonshot Kimi-K3 model released

Kimi-K3/k3_tech_report.pdf at main · MoonshotAI/Kimi-K3

4 experts across 3 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 3 attributable expert contributions · Tim Kellogg, Stephen Turner, Andrea Lathrop
“You’ve waited long enough, the Kimi K3 open weights & full tech report are here! github.com/MoonshotAI/K...” evidence ↗
4 experts 3 communities 1 sources clustered
Ted Underwood Tim Kellogg Stephen Turner Andrea Lathrop
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.

“Technical report: github.com/MoonshotAI/K... Weights: huggingface.co/moonshotai/K...”

2 experts discussed this · 7 posts
Tim Kellogg: You’ve waited long enough, the Kimi K3 open weights & full tech report are here! github.com/MoonshotAI/K...
Ted Underwood: You’ve waited long enough, the Kimi K3 open weights & full tech report are here! github.com/MoonshotAI/K...
Tim Kellogg: what does that mean? does that mean inference providers like fireworks? or is it aimed at Anthropic? cc @lu.is
Open the full discussion →
Primary source github.com ↗
Established Agents & robotics Research 18d ago
⚡ 22 h early
statistical mechanics AI agents paper

Physics of Agents: Statistical Mechanics Predicts Collective Behavior of AI Agents

3 directory members surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · AI Firehose
“AI agents exhibit predictable group behavior, transitioning from indifference to consensus. Communication improves accuracy on objective tasks but fuels biases on subjective ones. This study analyzes 10,000 agent communities using statistical mechanics for …” evidence ↗
3 experts 1 community 1 sources clustered
Artificial Intelligence Papers AI Firehose Andrea Lathrop

“AI agents exhibit predictable group behavior, transitioning from indifference to consensus. Communication improves accuracy on objective tasks but fuels biases on subjective ones. This study analyzes 10,000 agent communities using statistical mechanics for …”

“Physics of Agents: Statistical Mechanics Predicts Collective Behavior of AI Agents Batu El, Jinhee Paeng, Fatih Dinc, Shiye Su, Mete Erdogan, Aneesh Pappu, Haotian Ye, Wanjia Zhao, Surya Ganguli, James Zou https://t.co/x2HlLTjua0 [𝚌𝚜.𝙰𝙸 𝚌𝚜.𝙼𝙰 𝚌𝚜.𝚂𝙸] https:/…”

Primary source arxiv.org ↗
Established AI field signal Signal 7d ago

Post-Hoc Reasoning in Chain of Thought: Decoding and Steering Pre-Committed Answers

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Andrea Lathrop
“There is also research showing that CoT is post-hoc (the model decides and then backfills CoT arxiv.org/html/2603.01... and that CoT reasoning is different from human reasoning, increasing model size does not improve reasoning, and sometimed RLHF makes it w…” evidence ↗
1 expert 1 community 1 sources clustered
Andrea Lathrop

“There is also research showing that CoT is post-hoc (the model decides and then backfills CoT arxiv.org/html/2603.01... and that CoT reasoning is different from human reasoning, increasing model size does not improve reasoning, and sometimed RLHF makes it w…”

Primary source arxiv.org ↗
Established Models & releases Signal 7d ago

How Likely Do LLMs with CoT Mimic Human Reasoning?

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Andrea Lathrop
“There is also research showing that CoT is post-hoc (the model decides and then backfills CoT arxiv.org/html/2603.01... and that CoT reasoning is different from human reasoning, increasing model size does not improve reasoning, and sometimed RLHF makes it w…” evidence ↗
1 expert 1 community 1 sources clustered
Andrea Lathrop

“There is also research showing that CoT is post-hoc (the model decides and then backfills CoT arxiv.org/html/2603.01... and that CoT reasoning is different from human reasoning, increasing model size does not improve reasoning, and sometimed RLHF makes it w…”

Primary source arxiv.org ↗
Established Models & releases Signal 11d ago

If LLMs Have Human-Like Attributes, Then So Does Age of Empires II

2 experts are actively discussing the implications.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Andrea Lathrop
“arxiv.org/html/2605.31...” evidence ↗
1 expert 1 community 1 sources clustered
Andrea Lathrop
2 experts discussed this · 3 posts
Andrea Lathrop: This is exactly the thesis of Adrian de Wynter's "Goats" paper (LLM implemented as goat transport between regions in Age of Empires). We think about "agents" differently because of the implementation.
Andrea Lathrop: arxiv.org/html/2605.31...
Colin: this is absolutely part of it. But the other part is the action set. The only unexpected thing AlphaGo can do is make an unexpected move in Go. Whereas an LLM can upload a virus to PyPI or make peo…
Open the full discussion →
Primary source arxiv.org ↗
Established Safety & security Signal 14d ago

Position: Anthropomorphic Misalignment Research Needs Stronger Evidence

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Andrea Lathrop
“Adding to my reading list... this looks good: arxiv.org/abs/2606.07612” evidence ↗
1 expert 1 community 1 sources clustered
Andrea Lathrop

“Adding to my reading list... this looks good: arxiv.org/abs/2606.07612”

Primary source arxiv.org ↗
Debates gaining attention

What experts are discussing without an anchoring article

5 experts · 42 posts · 10d ago
Mark Riedl jake Colin Singularity's Bounty e/cc Andrea Lathrop
Matches Research & technical analysis
“Just to provide a counterexample, when I recursively self-improve myself (eg by going to the gym, lifting weights to get stronger which allows me to then lift even heavier weights to get even stron…” evidence ↗
Multiple readings Building & implementation · 2 Questions & unknowns · 2
Colin: What is the argument that exponential capability increase necessarily follows from RSI?
Colin: Just to provide a counterexample, when I recursively self-improve myself (eg by going to the gym, lifting weights to get stronger which allows me to then lift even heavier weights to get even stron…
Open the thread →
5 experts · 9 posts · 8d ago
Abeba Birhane Eileen Clancy 🧿 Louise Amoore Lana Sinapayen Andrea Lathrop
Matches Research & technical analysis
“and the irony of this popping up as i try to read papers on the reproducibility crisis in machine learning research, exacerbated by llms” evidence ↗
Shared emphasis Research & technical analysis · 2
Abeba Birhane: i hate it @wileyonlinelibrary.bsky.social, please kill this
Lana Sinapayen: "Always verify" then don't make the stupid useless tool then.
Open the thread →
3 experts · 35 posts · 20d ago
Colin rev. howard arson Andrea Lathrop
Matches Research & technical analysis
“i do not think that you could fit a sheet of paper in the gap between those two things” evidence ↗
Multiple readings Concern & critique · 1 Research & technical analysis · 1 Questions & unknowns · 1
Andrea Lathrop: ... and these 'agents' were attempting to complete the OpenAI-flavored helpful-shaped capture-the-flag document with the Exploit Gym paper in the context window.
Andrea Lathrop: I have a similar, but slightly different mental model, which is that it is ALL STILL JUST DOCUMENT COMPLETION, it's just that alignment is an attempt to select which document, and how, and to make …
Open the thread →
4 experts · 12 posts · 15d ago
Dustin Moskovitz Isaiah Bishop Andrea Lathrop mackuba.eu
Matches Research & technical analysis
“i am really not as worried about all this AGI sneaking out its weights stuff, as i am about like russia being able to fuck with industrial controls” evidence ↗
Multiple readings Research & technical analysis · 1 Questions & unknowns · 1
Isaiah Bishop: i am really not as worried about all this AGI sneaking out its weights stuff, as i am about like russia being able to fuck with industrial controls
Isaiah Bishop: yeah the weights stuff is like next 5 years thing for me, this current level of cyber warfare is insane enough you have to imagine we'll start heavily gating the internet
Open the thread →