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 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 Policy & governance
1 live reaction map in this view Jump to the evidence ↓
Multiple readings Unlawful by design: Exposing the human rights costs of generative AI - Amnesty International Policy & governance · 3 attributable reactions
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 Analysis 1d ago
⚡ 483 h early
Amnesty International generative AI human rights report

Unlawful by design: Exposing the human rights costs of generative AI - Amnesty International

11 experts across 2 network communities independently surfaced this.

Why this matches Concern & critique reaction 1 attributable expert contribution · Baldur Bjarnason
“> Without significant changes to how these systems are developed and deployed, they will continue to undermine international human rights laws and standards. www. amnesty.org/en/documents/pol40/0996/2026/en/ This is a pretty good report overall. Hard to dis…” evidence ↗
11 experts 2 communities 1 sources clustered
Iris van Rooij Luke Stark James Grimmelmann Alix Dunn tante rev. howard arson Pavel Adverb +3
How the network is reacting Experts are approaching this through 2 distinct lenses.
Multiple readings

Policy & governance

2 experts

Rules, institutions and accountability.

“"Amnesty International finds that standalone generative AI systems, based on unlawful web scraping, depend on mass invasions of privacy by design, and are fundamentally incompatible with international human rights law."”

Concern & critique

1 expert

Risks, limits and unintended consequences.

“> Without significant changes to how these systems are developed and deployed, they will continue to undermine international human rights laws and standards. www. amnesty.org/en/documents/pol40/0996/2026/en/ This is a pretty good report overall. Hard to dis…”

2 experts discussed this · 6 posts
James Grimmelmann: I know people who do a shitload of web scraping without the consent of the people whose servers and personal data are being scraped. And do you know why they do it? To document human rights violati…
James Grimmelmann: I’m sorry, but “unlawful web scraping” is not itself a human rights violation unless you have a view of either property rights or privacy rights that is itself incompatible with the human right to …
Adverb: I’m sorry, but “unlawful web scraping” is not itself a human rights violation unless you have a view of either property rights or privacy rights that is itself incompatible with the human right to …
Open the full discussion →
Primary source amnesty.org ↗
New Models & releases Signal 3h ago

Ai anthropic rsi recursive self improvement bff4e815

1 directory member surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · Nick Diakopoulos
“Here’s some Anthropic PR filtered through the WSJ: www.wsj.com/tech/ai/ai-a... — forget that we can’t actually evaluate the methods and biases in the data, or that the denominator of trivial tasks not requiring human judgment is likely to be dominated by AI.” evidence ↗
1 expert 1 community 1 sources clustered
Nick Diakopoulos

“Here’s some Anthropic PR filtered through the WSJ: www.wsj.com/tech/ai/ai-a... — forget that we can’t actually evaluate the methods and biases in the data, or that the denominator of trivial tasks not requiring human judgment is likely to be dominated by AI.”

Primary source wsj.com ↗
New Models & releases Signal 3h ago

Ai anthropic rsi recursive self improvement bff4e815

1 directory member surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · Nick Diakopoulos
“Here’s some Anthropic PR filtered through the WSJ: www.wsj.com/tech/ai/ai-a... — forget that we can’t actually evaluate the methods and biases in the data, or that the denominator of trivial tasks not requiring human judgment is likely to be dominated by AI.” evidence ↗
1 expert 1 community 1 sources clustered
Nick Diakopoulos

“Here’s some Anthropic PR filtered through the WSJ: www.wsj.com/tech/ai/ai-a... — forget that we can’t actually evaluate the methods and biases in the data, or that the denominator of trivial tasks not requiring human judgment is likely to be dominated by AI.”

Primary source wsj.com ↗
New Models & releases Development 16h ago
Claude restricts game recreation policy

Claude Can No Longer Produce A 'Recreation' Of Existing Games

1 directory member surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · James Grimmelmann
“Like every copyright-related AI guardrail, this will both prevent many legal uses and fail to prevent many infringing ones. Copyright is tough!” evidence ↗
1 expert 1 community 1 sources clustered
James Grimmelmann

“Like every copyright-related AI guardrail, this will both prevent many legal uses and fail to prevent many infringing ones. Copyright is tough!”

Primary source kotaku.com ↗
Developing Models & releases Research 1d ago
LLM bouba kiki preferences study

Bouba Beats Kiki, and Other Regularities in the Subjective Preferences of LLMs

1 directory member surfaced this signal.

Why this matches Concern & critique reaction 1 attributable expert contribution · AI Firehose
“Research shows LLMs exhibit distinct preferences in subjective decision-making tasks. Prompt variations influence these biases, challenging conventional evaluation methods and revealing the link between ideation and pruning in LLM outputs. https://arxiv.org…” evidence ↗
1 expert 1 community 1 sources clustered
AI Firehose

“Research shows LLMs exhibit distinct preferences in subjective decision-making tasks. Prompt variations influence these biases, challenging conventional evaluation methods and revealing the link between ideation and pruning in LLM outputs. https://arxiv.org…”

Primary source arxiv.org ↗
Debates gaining attention

What experts are discussing without an anchoring article

6 experts · 30 posts · 1d ago
Mar Hicks e.w. niedermeyer VE, cybersocial occult investigator Anthony Moser tweety fish Amy Hoy
Matches Concern & critique
“@vortexegg.com has described a spectrum of "slop vs care" I find useful because it captures how most ai/LLM related problems aren't really new, but existing harms amplified in that context we can t…” evidence ↗
Multiple readings Questions & unknowns · 2 Concern & critique · 1 Research & technical analysis · 1
e.w. niedermeyer: AI slop is a supply side game-changer, so it's easy to focus there, but ultimately everything comes down to demand: will the public, in aggregate, be satisfied with slop or demand more? skeptics ha…
e.w. niedermeyer: covering cars you have to separate your personal taste from what the market demands, and that's hard when it comes down to something like slop restaurant menus... I find them viscerally off-putting…
Open the thread →