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.

Concern & critique Research & technical analysis Context & explanation
2 live reaction maps in this view Jump to the evidence ↓
Reaction converging An alignment assessment of recent cybersecurity incidents Research & technical analysis · 7 attributable reactions Reaction converging Navier stokes solution Research & technical analysis · 6 attributable reactions
Showing signals surfaced by Colin ×
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 Signal 12d ago
⚡ 2 h early

An alignment assessment of recent cybersecurity incidents

14 experts across 5 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 6 attributable expert contributions · Ethan Mollick, Alexander Doria, Tim Kellogg
“Another incident of models escaping containment during a security test, this time Mythos 5. Lots going on here from a quick read. www.anthropic.com/research/ali...” evidence ↗
14 experts 5 communities 1 sources clustered
Ted Underwood Ethan Mollick Alexander Doria Tim Kellogg Ryan Moulton Andrew Arruda Rohan Paul Grace +6
4 experts discussed this · 15 posts
tweety fish: their goal is to have "an intelligence" which is "aligned" to doing things that are prosocial. They don't want to put in rules that STOP it from doing things: they see it fundamentally teleological…
tweety fish: one of the things I appreciate about this thread is that I don't think you can really understand the infosec stuff without understanding the ideological commitments the people making these have; li…
Ben Recht: yeah, now we're cooking...
Open the full discussion →
Primary source anthropic.com ↗
Established AI field signal Signal 19d ago
⚡ 7 h early

Navier stokes solution

14 experts across 6 network communities independently surfaced this.

Why this matches Research & technical analysis reaction 4 attributable expert contributions · Emad, Tim Kellogg, Alex Hern
“AI moves so fast Worth noting that the new model that showed this jump in capabilities has only been training since August 28th per the report and is now taking down problems in minutes Math (and hopefully soon physics) is getting bitter lesson'd https://t.…” evidence ↗
14 experts 6 communities 1 sources clustered
Emad Ethan Mollick Tim Kellogg Scott McGrath David Picard Alex Hern Sean O hEigeartaigh Dare Obasanjo +6
3 experts discussed this · 12 posts
Colin: Drama and palace intrigue aside, can anyone interpret what all of this implies about the status of Navier-Stokes? I don’t understand what anyone is saying.
Colin: It seems to me that where we stand is OpenAI is sitting on a 100 page PDF that claims to have solved (a version of) the Navier-Stokes problem and no person in the world knows if it's correct or not.
Colin: - OpenAI claims to have more-or-less done this tedious technical grinding to produce a resolution to NS proper, at least one version of it. - which they bizarrely offered authorship of this to Buck…
Open the full discussion →
Primary source openai.com ↗
Established Agents & robotics Resource 6d ago
Vending-Bench 2 agent benchmark

Vending-Bench 2 | Andon Labs

3 experts are actively discussing the implications.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Colin
“I think that Andon Labs' Vending-Bench is a nice illustration of the "just don't have that dog in them" theory of LLMs. This is a benchmark where an LLM agent is tasked with running a simulated vending machine, and scored on their earnings. GPT-6 Astra rece…” evidence ↗
1 expert 1 community 1 sources clustered
Colin

“I think that Andon Labs' Vending-Bench is a nice illustration of the "just don't have that dog in them" theory of LLMs. This is a benchmark where an LLM agent is tasked with running a simulated vending machine, and scored on their earnings. GPT-6 Astra rece…”

3 experts discussed this · 21 posts
Colin: Often people say "what's the difference?" but I think that this example makes the difference clear. A vending-machine maximizer with that dog in it would break the simulation and make a million dol…
Colin: It only ever wants to do one thing: write a story about the little guy that it's been trained to write a story about. That guy says a bunch of things that look like the kind of things that someone …
Colin: I think the answer is that telling the model "your goal is to maximize profits in this game" does not make its goal to maximize profits in this game. I'm positive there's a way to break the game an…
Open the full discussion →
Primary source andonlabs.com ↗
Established AI field signal Signal 5d ago

When the Machines Coordinate

7 experts are actively discussing the implications.

Why this matches Research & technical analysis reaction 1 attributable expert contribution · Colin
“there is one report that alleges this but I can't tell where they got this information from labs.cloudsecurityalliance.org/research/csa...” evidence ↗
1 expert 1 community 1 sources clustered
Colin

“there is one report that alleges this but I can't tell where they got this information from labs.cloudsecurityalliance.org/research/csa...”

7 experts discussed this · 78 posts
Singularity's Bounty e/cc: These are opening speeches from product liability lawyers in court 🤣 They don't want to explain, they want to litigate No, that's not right, prosecute
Colin: curious what you think of mine
Colin: If they had created a machine gun that fires in random directions and left it running as a test would I have to explain exactly how machine guns work?
Open the full discussion →
Primary source labs.cloudsecurityalliance.org ↗
Debates gaining attention

What experts are discussing without an anchoring article

6 experts · 78 posts · 4d ago
Henry Farrell Dustin Moskovitz Colin rev. howard arson tweety fish Matt Darling
Matches Research & technical analysis
“"the moral status of all counterfactual worldlines of positive measure forward of the present moment is identical to that or the one which actually occurs."” evidence ↗
Multiple readings Concern & critique · 1 Policy & governance · 1 Research & technical analysis · 1
rev. howard arson: what i am trying to get at here is that rationalism was an attempt to outline what turned out to be the extraordinary power of statistical systems at tasks for which they naively seem unsuited. vie…
rev. howard arson: "how do we get human values into these things" was broadly considered an intractable problem, and this was a central question of the movement even after we more or less made the first candidates of…
Open the thread →
8 experts · 36 posts · 3d ago
Dylan Hadfield-Menell Grace Colin Max Woolf austin Singularity's Bounty e/cc coty.bsky.social austegard.com
Matches Research & technical analysis
“I honestly think it would understand it okay, it might do better with longer context, I’m not sure. But I would definitely need to experiment with different ways of formatting the context” evidence ↗
Reaction converging Questions & unknowns · 3 Concern & critique · 1 Context & explanation · 1
Grace: I made a little Jev-based language model! It uses beam search + Jev as a classifier to judge branches based on coherence and "success" (whatever that means). I've barely talked to it so I don't kno…
Grace: (you'll need to tag him)
Open the thread →
6 experts · 38 posts · 4d ago
Thomas Dietterich Ciara Rowles Colin Mary Branscombe Singularity's Bounty e/cc Nafnlaus 🇮🇸 🇺🇦
Matches Research & technical analysis
“Ironically, researchers prefer to use "speech recognition" rather than "speech understanding" because the "understanding" usually overstates what the computer is doing.” evidence ↗
Multiple readings Concern & critique · 1 Research & technical analysis · 1 Opportunity & adoption · 1
Mary Branscombe: I don’t see the need to change AI since we’ve used that for the field since the 50s but yes, most of the others better reflect the actual capabilities and these are the kind of choices I already tr…
Mary Branscombe: Although models don’t just reflect bias, they both encode it and amplify and transfer it from one domain to another so I’d probably use exhibit
Open the thread →
5 experts · 53 posts · 5d ago
Naomi Saphra Ryan Moulton Matthew Kay Colin david-p-reichert.bsky.social
Matches Research & technical analysis
“You gotta tell Bender! (Though as Colin mentions in the other comment, maybe that's too stringent of a benchmark)” evidence ↗
Multiple readings Research & technical analysis · 2 Concern & critique · 1 Context & explanation · 1
Colin: OpenAI created autonomous hacking software and left it running for hours and days with no one watching
Colin: I’m don’t even care that much about the anthropomorphization thing, I think it can be fine in the right context, but this is perfectly succinct and understandable. If you want to go deeper into the…
Open the thread →