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 Research & technical analysis
1 live reaction map in this view Jump to the evidence ↓
Reaction converging IN THE WEIGHTS Research & technical analysis · 4 attributable reactions
Showing signals surfaced by austin ×
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 17d ago
⚡ 31 h early
In the Weights newsletter issue

IN THE WEIGHTS

12 experts across 5 network communities independently surfaced this.

Why this matches Concern & critique reaction 1 attributable expert contribution · Dare Obasanjo
“LLMs never fail to simultaneously impress and disappoint. Two searches for my name at intheweights.com showing completely different results with one accurate and one containing hallucinations.” evidence ↗
12 experts 5 communities 1 sources clustered
Lynn Cherny Kashmir Hill Sasha Costanza-Chock Reuben Binns Dare Obasanjo Ben Williamson Meredith Martin Aram Sinnreich +4
How the network is reacting 3 experts are independently emphasizing research & technical analysis.
Reaction converging

Research & technical analysis

3 experts

Evidence, methods and technical implications.

“are you in the weights? Hilarious and fascinating including the mess ups intheweights.com”

Concern & critique

1 expert

Risks, limits and unintended consequences.

“LLMs never fail to simultaneously impress and disappoint. Two searches for my name at intheweights.com showing completely different results with one accurate and one containing hallucinations.”

4 experts discussed this · 6 posts
austin: fun? I am in there. intheweights.com
VE, cybersocial occult investigator: Well, my strategy of having an incredibly common wallet name is still rendering me invisible to information systems.
austin: weirdly enough my github handle actually had a better score than my wallet name
Open the full discussion →
Primary source intheweights.com ↗
More conversations

What experts are discussing without an anchoring article

15 experts · 36 posts · 11h ago
Mike Masnick Scott McGrath Grace Liz Fong-Jones (方禮真) jake Dan Hon Doll Alex Rubinsteyn
Matches Concern & critique
“Did you try this yourself? Because I feel like a lot of people who dismiss AI coding outright have no idea what it's like. People are criticizing and fighting the idea of it they have in their head…” evidence ↗
Multiple readings Concern & critique · 3 Research & technical analysis · 2 Policy & governance · 1
Grace: He's upfront with when he uses AI, so you can trust he's upfront with when he doesn't
mackuba.eu: Please just stop…
Open the thread →
11 experts · 78 posts · 14d ago
Mike Masnick SE Gyges Liz Fong-Jones (方禮真) jake Arseny Khakhalin Doll austin Alex Gude
Matches Concern & critique
“Criticism is ok, but what a lot of people are doing here is not criticism (some are openly bragging that they're doing bullying and will be bullying even more)” evidence ↗
Reaction converging Opportunity & adoption · 3 Concern & critique · 1 Questions & unknowns · 1
mackuba.eu: FFS, this is so fucked… 😩 Would it kill the people here to not chase away any tech person who comes here and says something about AI?… I really want tech discussions and communities to hang out her…
Mike Masnick: And now I'm getting shit upon for merely repeating what people told me, which some jackasses are misrepresenting. 🙄 Doing a lot of blocking unfortunately.
Open the thread →
3 experts · 58 posts · 13h ago
Liz Fong-Jones (方禮真) Doll austin
Matches Concern & critique
“at $200 a month (plus, let's say $100-400 a month in overages), it was absolutely a steal, zero regrets. at $6k (10-20x the cost), it is pretty eye-wateringly expensive. anthropic is phasing out th…” evidence ↗
Liz Fong-Jones (方禮真): for an engineer like me who is maximally using Sonnet, and hasn't yet switched to a local model, it's about $6k a month, or $72k a year. it is a very very hefty fraction of the salary, & is going t…
Liz Fong-Jones (方禮真): this is definitely not technology that, at the moment, you want put in the hands of the average white-collar worker right now, you'd be better off just hiring more workers, rather than pay for the …
Open the thread →
4 experts · 26 posts · 8d ago
Hypervisible VE, cybersocial occult investigator austin Teresa Heffernan
Matches Concern & critique
“It’s not just that guardrails are “advisory” in nature — guardrails have exactly the same bargain incantation to the bullshit engine nature as all forms of “direction” to models do. A human-biased …” evidence ↗
Shared emphasis Building & implementation · 2
VE, cybersocial occult investigator: Something I notice watching top trending github repos and the eng discourse on bsky (on an alt acct) is that a significant amount of the popularly socialized software engineering behavior now is fu…
VE, cybersocial occult investigator: That’s pretty much what a lot of skills, agent frameworks, harnesses, etc amount to at the end of the day. And it’s not that these sort of artifacts do nothing at all — I have enough hands on exper…
Open the thread →
3 experts · 41 posts · 15d ago
Colin austin tj mahr 🤘
Matches Concern & critique
“Please don't get me wrong. People Who Use AI should not be a protected class. Saying dehumanizing things about people who use AI is not the same kind of thing as saying something racist or homophob…” evidence ↗
Colin: You're going to think this is a reach but I think that trypophobia burrito is a good example of a kind of subtle dehumanization that's hiding inside of a lot of ragebait. The image is disgusting. I…
Colin: The feeling of disgust at something that appears to be rotting, maggot-filled, etc., is a universal human experience. Anyone who looks at the trypophobia burrito can see that it's disgusting. What …
Open the thread →
3 experts · 12 posts · 6d ago
VE, cybersocial occult investigator austin Abstract Tesseract
Matches Concern & critique
“I haven’t really found any of those yet… the city I live in has a tech scene but it’s become dominated military robotics and milgov surveillance AI companies” evidence ↗
VE, cybersocial occult investigator: I had lately found my anger about technofascism to be so much that it was making it hard for me to feel in community with other tech people—just in general, not even ones who support fascism—so see…
Abstract Tesseract: yeah I get that it's not sustainable for everyone to be angry about everything all the time, but given how much these communities have historically been willing to overlook, it does make me feel be…
Open the thread →
2 experts · 20 posts · 4d ago
VE, cybersocial occult investigator austin
Matches Concern & critique
“I think the frame we should really be exploring is that the world is suffering from a crisis of _what the software is actually being produced to do_; such as fascist surveillance, behavioral and ps…” evidence ↗
Shared emphasis Concern & critique · 2
VE, cybersocial occult investigator: I think the frame we should really be exploring is that the world is suffering from a crisis of _what the software is actually being produced to do_; such as fascist surveillance, behavioral and ps…
VE, cybersocial occult investigator: This is entirely orthogonal issue to both the quantity and quality of software produced, and actually driving up either the quantity of software production OR the quality of software production mak…
Open the thread →
4 experts · 7 posts · 11d ago
Dan Hon Doll austin mackuba.eu
Matches Concern & critique
“Yeah nobody wants their IP to get logged even if ant super promises not to steal it. (because they will)” evidence ↗
Multiple readings Concern & critique · 1 Building & implementation · 1
mackuba.eu: Yeah, OpenAI launched 5.6 as "Sol is the best, but Terra is for normal everyday stuff", but in practice it's totally fine for me to just have Sol set permanently and use it for everything
Dan Hon: Claude make me an artifact that illustrates price elasticity
Open the thread →
2 experts · 23 posts · 15d ago
Phillip Carter austin
Matches Concern & critique
“there’s a few takeaways here imo: 1. probabilistic systems are gonna fail! they’re gonna have errors and make mistakes. build resilient systems.” evidence ↗
austin: saw someone say AI can’t write SQL some data for anyone that cares - @honeycomb.io offers agents a query tool against our datastore; it’s a custom JSON schema, which means it’s very unlikely to be …
austin: the upshot of this is that the tool basically gives the agent some docs on how to use it, and the schema itself. maybe 3 examples, but a lot of the “how do you use this” comes from interpreting the…
Open the thread →
2 experts · 11 posts · 3d ago
SE Gyges austin
Matches Concern & critique
“imo it's less that we're stuck with bash and more that the alternatives are worse in several ways. I think you can see it with OAI responses/harmony stuff happening in real-time -- you can make it …” evidence ↗
SE Gyges: i am and always have been a text absolutist, and i think that the model's interface to the computer should be the terminal and only the terminal. every model-specific connector is an admission that…
SE Gyges: that is: in terms of what an LLM that handles text does, it should be given a fucking terminal. not an mcp. not an endless array of tools. a terminal. we have a tool for calling external services f…
Open the thread →
3 experts · 7 posts · 5d ago
tweety fish austin Matt Darling
Matches Concern & critique
“human narrative chain of thought is not (or at least, is not obligatorily) on the critical path for decision-making tho, it's more pure after-the-fact narration.” evidence ↗
austin: human narration of their own chain of thought is also not reliable 🫨
Matt Darling: human narration of their own chain of thought is also not reliable 🫨
Open the thread →
3 experts · 6 posts · 5d ago
austin Singularity's Bounty e/cc david-p-reichert.bsky.social
Matches Concern & critique
“... we might get global collaboration on AI, but I doubt it'll be because of copyright). IMO the more promising route is to focus on the downstream harms & benefits of the tech. We all contributed …” evidence ↗
Multiple readings Opportunity & adoption · 1 Questions & unknowns · 1
austin: it may shock you to learn that people do not really appreciate a) the distinction or b) that there is no version of AI that precludes ‘the bad stuff’
david-p-reichert.bsky.social: Agree not everything should be called AI, but on your second point, fwiw AGI is the one with general in it and that's still around, I'm not aware of gen(erative) AI being a rebranding..
Open the thread →