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 Questions & unknowns
3 live reaction maps in this view Jump to the evidence ↓
Multiple readings Hugging face model evaluation security incident Concern & critique · 5 attributable reactions Multiple readings The System From Nowhere Concern & critique · 3 attributable reactions Reaction converging Elias in the Lighthouse, Again? Diagnosing Low Diversity in LLM Stories Research & technical analysis · 5 attributable reactions
Showing signals surfaced by Eryk Salvaggio ×
Debated

The questions experts are actively pulling apart

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

Established Evaluation & benchmarks Signal 13d ago
⚡ 1 h early

Hugging face model evaluation security incident

22 experts across 6 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Jeffrey P. Bigham
“model capability jaggedness is part of the unintuitiveness of current AI, but it's made even less intuitive by tirelessness… wigguming through the jaggedness toward something that looks like success. not quite a paperclip factory, but not so far off. metaph…” evidence ↗
22 experts 6 communities 1 sources clustered
Ethan Mollick Thomas Steinke Jeffrey P. Bigham Eryk Salvaggio Deedy Das Mor Naaman Tim Kellogg Djamé 🟥 +14
6 experts discussed this · 11 posts
Grace: This headline is extremely funny given what happened (OpenAI hacked HF by accident) openai.com/index/huggin...
Grace: Maybe the most concerning part is the OpenAI claim to not have known about this before investigating?
Grace: Well, I think the model passed the test
Open the full discussion →
Primary source openai.com ↗
Developing Compute & infrastructure Analysis 2d ago
⚡ 6 h early
system from nowhere essay

The System From Nowhere

6 experts across 2 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Eryk Salvaggio
“The System From Nowhere is a media problem: What (the system) is optimized to do; what training data shapes it; what counts as a valid response: these are mechanisms put in place by human decisions. Good reporting on AI would discuss them. Instead, they “no…” evidence ↗
6 experts 2 communities 1 sources clustered
Eryk Salvaggio Eileen Clancy 🧿 Dr Heidy Khlaaf (هايدي خلاف) Frank Pasquale Neil Turkewitz Jeffrey Vagle
How the network is reacting Experts are approaching this through 3 distinct lenses.
Multiple readings

Concern & critique

1 expert

Risks, limits and unintended consequences.

““When we say “AI models went rogue,” we skip the entire story: the part where OpenAI manually removed the model's cybersecurity blocks. We skip that OpenAI chose to test it on a machine with a live network connection.” mail.cyberneticforests.com/the-system-…”

Research & technical analysis

1 expert

Evidence, methods and technical implications.

“The System From Nowhere is a media problem: What (the system) is optimized to do; what training data shapes it; what counts as a valid response: these are mechanisms put in place by human decisions. Good reporting on AI would discuss them. Instead, they “no…”

Questions & unknowns

1 expert

What remains unresolved or contested.

“"We're told that AI did something, and journalists don't have to wade into why or how. It lets them cover a story without raising technically complicated questions...or confusing questions about the way they’re built and why they are built that way." mail.c…”

2 experts discussed this · 6 posts
Eryk Salvaggio: “Nobody is losing control of an emerging intelligence. They’re redirecting their own agency to the thing they are building.” It’s a myth, a self-affirming form of sense making: they take the action…
Eryk Salvaggio: The myth erases the origins of these models — reframing them as beyond human control or design — but a point of view accumulates within the model from the decisions engineers make about it. The myt…
Dr Heidy Khlaaf (هايدي خلاف): “Nobody is losing control of an emerging intelligence. They’re redirecting their own agency to the thing they are building.” It’s a myth, a self-affirming form of sense making: they take the action…
Open the full discussion →
Primary source mail.cyberneticforests.com ↗
Established AI research Research 8d ago
⚡ 43 h early
latent diffusion neural economy paper

The Market in the Model: Latent Diffusion as Neural Economy

3 experts across 2 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Eryk Salvaggio
“But my favorite nerdy observation is that Async is an MRI company — and the latent diffusion process is built on U-Net, which was widely used in medical imaging (I call this the move from corpora to corpus in my “market in the model” pre-print!) arxiv.org/a…” evidence ↗
3 experts 2 communities 1 sources clustered
Eryk Salvaggio Matthew Kirschenbaum Roland Meyer

“But my favorite nerdy observation is that Async is an MRI company — and the latent diffusion process is built on U-Net, which was widely used in medical imaging (I call this the move from corpora to corpus in my “market in the model” pre-print!) arxiv.org/a…”

2 experts discussed this · 8 posts
Eryk Salvaggio: The way these models work now, any mechanized “taste” (aesthetic preferences) operates as a constraints on what models produce; that constraint would be recognizable in images people shared. They w…
Eryk Salvaggio: Taste, as a curatorial instinct that balances novelty and familiarity, is already re-enacted by these models. We don’t recognize it as such but these mechanisms aim to “automate discernment;” LoRAs…
Eryk Salvaggio: Aesthetically valued diffusion-generated images exist, they just require calibrating the discernment mechanism away from default model aesthetics. That has to be rare to hold information. But look …
Open the full discussion →
Primary source arxiv.org ↗
Established Models & releases Research 6d ago
⚡ 17 h early
LLM story diversity diagnosis paper

Elias in the Lighthouse, Again? Diagnosing Low Diversity in LLM Stories

5 experts across 3 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 5 attributable expert contributions · Ted Underwood, David Mimno, Eryk Salvaggio
“In the "Limitations" section they acknowledge that their study was monolingual, but say a multilingual generalization would be interesting. arxiv.org/abs/2605.26492” evidence ↗
5 experts 3 communities 1 sources clustered
Ted Underwood David Mimno Eryk Salvaggio Grace Sil Hamilton
How the network is reacting 5 experts are independently emphasizing research & technical analysis.
Reaction converging

Research & technical analysis

5 experts

Evidence, methods and technical implications.

“In the "Limitations" section they acknowledge that their study was monolingual, but say a multilingual generalization would be interesting. arxiv.org/abs/2605.26492”

2 experts discussed this · 2 posts
Ted Underwood: was reading this and enjoying it — and then enjoyed it more when it turned out to be a paper by Sil Hamilton and @dmimno.bsky.social
David Mimno: was reading this and enjoying it — and then enjoyed it more when it turned out to be a paper by Sil Hamilton and @dmimno.bsky.social
David Mimno: The preprint is arxiv.org/abs/2605.26492, and @srhm.ca is working on a post!
Open the full discussion →
Primary source arxiv.org ↗
Established AI field signal Signal 16d ago

Evaluating Verifiability in Generative Search Engines

2 experts are actively discussing the implications.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Eryk Salvaggio
“For example, web search and summarization was terrible: In 2023 51.5% of Web summaries misstated what they cited. They were rolled out anyway (and sometimes rolled back). arxiv.org/abs/2304.09848” evidence ↗
1 expert 1 community 1 sources clustered
Eryk Salvaggio

“For example, web search and summarization was terrible: In 2023 51.5% of Web summaries misstated what they cited. They were rolled out anyway (and sometimes rolled back). arxiv.org/abs/2304.09848”

2 experts discussed this · 11 posts
Eryk Salvaggio: Much respect to Ted but we could just as well flip this. Three years ago companies were making claims that were genuinely inaccurate descriptions of their capabilities. Pure LLMs *were* terrible se…
Eryk Salvaggio: For example, web search and summarization was terrible: In 2023 51.5% of Web summaries misstated what they cited. They were rolled out anyway (and sometimes rolled back). arxiv.org/abs/2304.09848
Eryk Salvaggio: Remember the claim in 2023 that "Theory of Mind had spontaneously emerged" in GPT-3? Also in 2023 AutoGPT claimed it could automate agentic workflows and failed spectacularly. Many critics were jus…
Open the full discussion →
Primary source arxiv.org ↗