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 Building & implementation Policy & governance Concern & critique
4 live reaction maps in this view Jump to the evidence ↓
Reaction converging Research acceleration view inside openai Research & technical analysis · 6 attributable reactions Reaction converging Countering misuse of AI: September 2026 / Anthropic Concern & critique · 7 attributable reactions Multiple readings Anthropic ai biological weapons Building & implementation · 4 attributable reactions +1 more reaction maps further down
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.

New Models & releases Signal 9h ago
⚡ 56 h early

Research acceleration view inside openai

11 experts across 5 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 4 attributable expert contributions · Tim Kellogg, Boris Power, René Walter
“OpenAI has an automated AI “research intern” openai.com/index/resear...” evidence ↗
11 experts 5 communities 1 sources clustered
Ramon Astudillo Tim Kellogg Sung Kim Dare Obasanjo Boris Power Erik Brynjolfsson Elie Ricardo Barros Lourenço +3
Primary source openai.com ↗
New Models & releases Development 8h ago
⚡ 1 h early
Anthropic AI misuse September report

Countering misuse of AI: September 2026 / Anthropic

15 experts across 4 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Michiel Bakker
“@kotekjedi_ml Here Ant's report https://t.co/22vHruWVqH And the paper from @kotekjedi_ml @DavidSchmotz @iliaishacked https://t.co/wuCpNZaQ0d” evidence ↗
15 experts 4 communities 1 sources clustered
Jason Koebler Tim Kellogg Michiel Bakker Andrew Arruda John Burn-Murdoch Kevin S. Xu Eileen Clancy 🧿 Martin Trapp +7
How the network is reacting 5 experts are independently emphasizing concern & critique.
Reaction converging

Concern & critique

5 experts

Risks, limits and unintended consequences.

“will have an article about ai agents some point next week but where i'm at rn is we are seeing AI get access to accounts/systems/the internet en masse. as directed by humans, that is gonna cause big problems anthropic threat report yesterday was interesting…”

Building & implementation

1 expert

How teams are shipping and applying it.

“Thomson Reuters announced it is moving off Claude to Alibaba's Qwen to cut costs. From today's Anthropic report: Alibaba extracted 151M+ Claude exchanges to help train Qwen. TR now deploys Westlaw on a vast trove of stolen US IP. @AnthropicAI's report: http…”

Research & technical analysis

1 expert

Evidence, methods and technical implications.

“@kotekjedi_ml Here Ant's report https://t.co/22vHruWVqH And the paper from @kotekjedi_ml @DavidSchmotz @iliaishacked https://t.co/wuCpNZaQ0d”

Primary source anthropic.com ↗
New Models & releases Signal 7h ago

Anthropic ai biological weapons

8 experts across 4 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Shannon Mattern
““Anthropic said it could not determine whether the research served a legitimate or nefarious purpose because valid biological inquiry… can also help engineer dangerous pathogens.” Reminiscing on the days when we had peer review panels with enough subject ex…” evidence ↗
8 experts 4 communities 1 sources clustered
Justin Hendrix Ryan Mac 🙃 Shannon Mattern Anna Mills Eileen Clancy 🧿 Alejandra Caraballo Emily Feng Stephen Turner
Primary source nytimes.com ↗
Accelerating AI business Signal 6h ago

Anthropic ceo dario amodei calls ai industry slow down

3 experts across 3 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Berna Devezer
“ivy full prof who publishes 25 papers a year calls for academics to publish less” evidence ↗
3 experts 3 communities 1 sources clustered
Karl Bode Berna Devezer Jess Miers 🦝🦞
Primary source washingtonpost.com ↗
New Models & releases Research 15h ago
⚡ 101 h early
LLM war judgment alignment paper

Language models judge war differently when tested for alignment

2 directory members surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · John Horton
“Interesting paper from @maxchupilkin illustrating a kind of 'Volkswagen emissions test" effect where LLMs respond differently about war when being told then are being tested for alignment https://t.co/rMWObcyPZI 1/ https://t.co/g9rysSZk3p” evidence ↗
2 experts 0 communities 1 sources clustered
John Horton Artificial Intelligence Papers

“Interesting paper from @maxchupilkin illustrating a kind of 'Volkswagen emissions test" effect where LLMs respond differently about war when being told then are being tested for alignment https://t.co/rMWObcyPZI 1/ https://t.co/g9rysSZk3p”

“Language models judge war differently when tested for alignment Maxim Chupilkin https://t.co/YfyoMfeA08 [𝚌𝚜.𝙰𝙸 𝚌𝚜.𝙲𝚈] https://t.co/0voM3qtf8M”

Primary source arxiv.org ↗
New Models & releases Research 15h ago
⚡ 20 h early
clinical LLM deterministic math solver

Towards a Deterministic Math Solver for Clinical Language Models

2 directory members surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · AI Firehose
“MIT researchers introduced a method to enhance clinical calculators, allowing models to produce Python code for calculations. The 'Program-Solve' approach yields better accuracy in medical computations, surpassing standard arithmetic. https://arxiv.org/abs/…” evidence ↗
2 experts 1 community 1 sources clustered
Artificial Intelligence Papers AI Firehose

“MIT researchers introduced a method to enhance clinical calculators, allowing models to produce Python code for calculations. The 'Program-Solve' approach yields better accuracy in medical computations, surpassing standard arithmetic. https://arxiv.org/abs/…”

“Towards a Deterministic Math Solver for Clinical Language Models Felipe Ocampo Osorio, Sebastián Andrés Cajas Ordoñez, Maximin Lange, Rafi Al Attrach, Sahil Kapadia, Zakaria Laouabdia Sellami, Angelo Antonio Talio, Leo Anthony Celi https://t.co/aAT7Z0DMoe […”

Primary source arxiv.org ↗
New Models & releases Signal 6h ago
⚡ 3 h early

Holistic Evaluation of Language Models

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Christopher Manning
“Holistic Evaluation of Language Models – https://t.co/jEje8m0mDp ELEPHANT: Measuring and understanding social sycophancy in LLMs – https://t.co/sHYjxabKQx Sycophantic AI decreases prosocial intentions and promotes dependence – https://t.co/aTSrtInvKS AI gen…” evidence ↗
1 expert 1 community 1 sources clustered
Christopher Manning

“Holistic Evaluation of Language Models – https://t.co/jEje8m0mDp ELEPHANT: Measuring and understanding social sycophancy in LLMs – https://t.co/sHYjxabKQx Sycophantic AI decreases prosocial intentions and promotes dependence – https://t.co/aTSrtInvKS AI gen…”

Primary source arxiv.org ↗
New Models & releases Signal 6h ago
⚡ 3 h early

ELEPHANT: Measuring and understanding social sycophancy in LLMs

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Christopher Manning
“Holistic Evaluation of Language Models – https://t.co/jEje8m0mDp ELEPHANT: Measuring and understanding social sycophancy in LLMs – https://t.co/sHYjxabKQx Sycophantic AI decreases prosocial intentions and promotes dependence – https://t.co/aTSrtInvKS AI gen…” evidence ↗
1 expert 1 community 1 sources clustered
Christopher Manning

“Holistic Evaluation of Language Models – https://t.co/jEje8m0mDp ELEPHANT: Measuring and understanding social sycophancy in LLMs – https://t.co/sHYjxabKQx Sycophantic AI decreases prosocial intentions and promotes dependence – https://t.co/aTSrtInvKS AI gen…”

Primary source arxiv.org ↗
New Models & releases Signal 17h ago

Circuit Tracing: Revealing Computational Graphs in Language Models

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Alexander Doria
“Yes with CoT. Transformer circuit also has experiments with straight (simpler) operations. transformer-circuits.pub/2025/attribu...” evidence ↗
1 expert 1 community 1 sources clustered
Alexander Doria

“Yes with CoT. Transformer circuit also has experiments with straight (simpler) operations. transformer-circuits.pub/2025/attribu...”

Primary source transformer-circuits.pub ↗
New Models & releases Research 2h ago
diffusion VLM trajectory decoding paper

Routing by Reasoning Need: Trajectory-Aware Decoding Control for Diffusion Vision-Language Models

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Artificial Intelligence Papers
“Routing by Reasoning Need: Trajectory-Aware Decoding Control for Diffusion Vision-Language Models Yixiang Liu, Zhongxing Xu, Zhonghua Wang, Xiaoying Tang https://t.co/Y2OqHd88sG [𝚌𝚜.𝙰𝙸] 💬Accepted to Findings of EMNLP 2026 https://t.co/mA6gOnriLf” evidence ↗
1 expert 0 communities 1 sources clustered
Artificial Intelligence Papers

“Routing by Reasoning Need: Trajectory-Aware Decoding Control for Diffusion Vision-Language Models Yixiang Liu, Zhongxing Xu, Zhonghua Wang, Xiaoying Tang https://t.co/Y2OqHd88sG [𝚌𝚜.𝙰𝙸] 💬Accepted to Findings of EMNLP 2026 https://t.co/mA6gOnriLf”

Primary source arxiv.org ↗
New Evaluation & benchmarks Signal 8h ago

Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Artificial Intelligence Papers
“Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment Jiayu Huang, Zichen Tang, Qianhui Ling, Zemin Kuang, Haihong E https://t.co/pWSPJvr8SF [𝚌𝚜.𝙰𝙸] 💬Accepted to EMNLP 2026 Main Conference https://…” evidence ↗
1 expert 0 communities 1 sources clustered
Artificial Intelligence Papers

“Can LLMs Follow Medical Expert Logic? A Benchmark for Hierarchical Logical Consistency in Risk-of-Bias Assessment Jiayu Huang, Zichen Tang, Qianhui Ling, Zemin Kuang, Haihong E https://t.co/pWSPJvr8SF [𝚌𝚜.𝙰𝙸] 💬Accepted to EMNLP 2026 Main Conference https://…”

Primary source arxiv.org ↗
New Evaluation & benchmarks Signal 8h ago

SemVerBench: Benchmarking LLM Comprehension of Version-Constraint Resolution Semantics

1 directory member surfaced this signal.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Artificial Intelligence Papers
“SemVerBench: Benchmarking LLM Comprehension of Version-Constraint Resolution Semantics Qibai Chen, Zeming Liu https://t.co/jSncWC8Mel [𝚌𝚜.𝙰𝙸 𝚌𝚜.𝚂𝙴] 💬Accepted at the 38th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2026) https:…” evidence ↗
1 expert 0 communities 1 sources clustered
Artificial Intelligence Papers

“SemVerBench: Benchmarking LLM Comprehension of Version-Constraint Resolution Semantics Qibai Chen, Zeming Liu https://t.co/jSncWC8Mel [𝚌𝚜.𝙰𝙸 𝚌𝚜.𝚂𝙴] 💬Accepted at the 38th IEEE International Conference on Tools with Artificial Intelligence (ICTAI 2026) https:…”

Primary source arxiv.org ↗
Debates gaining attention

What experts are discussing without an anchoring article

2 experts · 8 posts · 1d ago
Berna Devezer tweety fish
Matches Research & technical analysis
“Engaging with a colleague, with whom you mostly communicate via email, becomes a chore when they only ever share pure slop. I'd rather fill out paperwork than actually spend gray matter and time on…” evidence ↗
Shared emphasis Research & technical analysis · 2
Berna Devezer: Totally vibing here, but hope we'll soon get collectively fed up with llm-speak and refuse to entertain it. Getting impossible to seriously engage with extensive text without an underlying thought …
Berna Devezer: Engaging with a colleague, with whom you mostly communicate via email, becomes a chore when they only ever share pure slop. I'd rather fill out paperwork than actually spend gray matter and time on…
Open the thread →