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
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.

Building & implementation Research & technical analysis
1 live reaction map in this view Jump to the evidence ↓
Multiple readings Zoom In: An Introduction to Circuits Building & implementation · 2 attributable reactions
Showing signals surfaced by jake ×
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 AI research Resource 12d ago
⚡ 1465 h early
Distill.pub circuits neural network introduction

Zoom In: An Introduction to Circuits

3 experts across 3 network communities independently surfaced this.

3 experts 3 communities 1 sources clustered
jake Mary Branscombe Nafnlaus 🇮🇸 🇺🇦
How the network is reacting Experts are approaching this through 2 distinct lenses.
Multiple readings

Building & implementation

1 expert

How teams are shipping and applying it.

“NNs in general build up, from the base level, like this: distill.pub/2020/circuit...”

Research & technical analysis

1 expert

Evidence, methods and technical implications.

“if you want to understand more about how weird the "using statistics to predict the next word" works out to in practice, you try the mechanistic interpretability work from anthropic: distill.pub/2020/circuit... neelnanda.io/mechanistic-... transformer-circu…”

4 experts discussed this · 37 posts
mackuba.eu: Many are contrasting the bsky response to harassment of trans users to the harassment of AI developers and users. FWIW, as a transgender developer and user of AI systems, anecdotally, I've received…
Olivia Guest: 💯
jake: lyInG tHeFT mAchIne thAt's rUInInG peOple'S wAtER
Open the full discussion →
Primary source distill.pub ↗
More conversations

What experts are discussing without an anchoring article

8 experts · 30 posts · 7d ago
Ted Underwood jake Colin tweety fish René Walter Matt Darling Singularity's Bounty e/cc Andrea Lathrop
Multiple readings Concern & critique · 1 Building & implementation · 1 Questions & unknowns · 1
Colin: Frankenmodel refuses to believe it's not just Claude
Colin: I don't think if you put 6 people in a circle and asked them to tell a story one word at a time they would, together, spontaneously create a story which denies that they are indeed six individuals
Open the thread →
5 experts · 27 posts · 19h ago
Mark J. Nelson Eli Sennesh SE Gyges Meredith Martin jake
Shared emphasis Questions & unknowns · 2
SE Gyges: unpopular opinion, "admits uncertainty" and "flags nuances" are shit. they are terrible fucking traits actually. is the sky blue? do we really know that? what is blue, anyway? many cultures have di…
SE Gyges: if gemini argon is a remarkably good model it is good because it is simply not optimized for this specific set of toxic, absurd, brain-poisoned behavior. it would actually defy belief that it happe…
Open the thread →
4 experts · 31 posts · 5d ago
Maxim Raginsky jake Colin tweety fish
Shared emphasis Research & technical analysis · 2
Colin: I see people say this all the time and I think it’s just wrong in general. I don’t think you can make human skin out of pebbles. This is not because I think human skin is magic and pebbles are non-…
Colin: Consciousness may or may not be the kind of thing you can make out of artificial neural networks. Magic does not need to have anything to do with it.
Open the thread →
4 experts · 14 posts · 20h ago
Eugene Vinitsky SE Gyges jake Doll
Multiple readings Research & technical analysis · 1 Questions & unknowns · 1
SE Gyges: "my best case future involves people competing viciously with each other for power forever" is just so fucking grim to me man
SE Gyges: if you are mad that we are not moving to the star trek future you should, like, try to chart course for that future. why would anyone do less
Open the thread →
3 experts · 16 posts · 12d ago
jake Doll Singularity's Bounty e/cc
jake: I think we’re reading a bit too much into what he’s saying here. He’s probably just not communicating his feelings properly here.
jake: He has written about how “unsolved problems that humanity cares about” are a resource which needs careful tending to, and I sympathize with that: human understanding *is* needed if we are to *care*.
Open the thread →
2 experts · 14 posts · 17d ago
Benj Edwards jake
Benj Edwards: Also, some people have the mistaken impression that AI agents aren't very capable, based on the shortcomings of AI models as recently as January. But capabilities have improved rapidly, and they ar…
Benj Edwards: We are in an early period where demand for compute is outstripping supply, and it's leading to a speculative build-out of datacenters fueled by Big Tech billions that are delivering more capable AI…
Open the thread →
3 experts · 3 posts · 6d ago
Eugene Vinitsky Chris Paxton jake
Multiple readings Research & technical analysis · 1 Questions & unknowns · 1
Chris Paxton: The solution to interpretability is clearly the same as the solution to ai coding: just ask deepseek if claude did what it said it did, they can't BOTH be lying to you
Eugene Vinitsky: Okay fine they both lied but if I ask 3 models it’s all good then we can do majority vote
Open the thread →
2 experts · 10 posts · 16d ago
Tim Kellogg jake
Tim Kellogg: poll: do you use embeddings? i’m finding a growing trend of people shedding embeddings for increasingly agentic flows i still think embeddings are useful, but not as broadly or generally as we used…
Tim Kellogg: i also think Jev will cut into even more embedding use cases like you could use embeddings for classification. Also, Jev makes feature extraction easier, so you can just use SQL and then you also d…
Open the thread →
2 experts · 6 posts · 11d ago
Tim Kellogg jake
Multiple readings Concern & critique · 1 Opportunity & adoption · 1
Tim Kellogg: second: if you’re a software engineer and you answered yes or yes-ish to this, do you feel threatened by AI?
jake: Nope. I think he was right, in the very detailed way of "all source code" (to a reasonable approximation) is now being written by LLMs. But that doesn't mean that all *software* is being built (exc…
Open the thread →