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.

2,364 searchable experts 2,966 tracked across all sources
Clear
Showing signals surfaced by Simon Willison ×
Debated

The questions experts are actively pulling apart

One card per development. Sources are clustered; expert reactions remain attributable.

Established AI field signal Resource 17d ago
⚡ 17 h early
AI Compass navigation guide resource

The AI Compass

15 experts across 5 network communities independently surfaced this.

15 experts 5 communities 1 sources clustered
Maria Antoniak Simon Willison Brian Merchant Scott McGrath Grace Nicole Hennig Dan Greene tante +7

“not wrong! bambamramfan.github.io/ai-compass/”

“Source code here: github.com/bambamramfan... - try the quiz here: bambamramfan.github.io/ai-compass/”

4 experts discussed this · 5 posts
Grace: Did anyone else get this?
David Picard: Patron saint: Gary Marcus? Ouch! 😬🫣
Tim Duffy: I do recall one question where I wanted there to be one more extreme answer past the final option
Open the full discussion →
Primary source bambamramfan.github.io ↗
Established AI field signal Analysis 19d ago
⚡ 3 h early
Fable AI company judgment

Fable's judgement

4 experts are actively discussing the implications.

2 experts 2 communities 1 sources clustered
Ethan Mollick Simon Willison

“Simon’s blog: simonwillison.net/2026/Jul/3/j...”

“The most interesting Fable tip I've heard so far is to let the model use its own judgement as much as possible I told it "For all coding tasks use your judgement to decide an appropriate lower power model and run that in a subagent" and it seems to be savin…”

4 experts discussed this · 6 posts
Ethan Mollick: What if the model is the router? I think people underestimate the ability of frontier models now, but especially in the near future, to delegate work on their own as needed to dumber, cheaper model…
Ethan Mollick: Simon’s blog: simonwillison.net/2026/Jul/3/j...
Simon Willison: That paper predates Fable - I'd love to see them re-run that experience to see if Fable is a meaningful improvement My hunch is that it does, but data is better then hunches!
Open the full discussion →
Primary source simonwillison.net ↗
Established Evaluation & benchmarks Analysis 7d ago
⚡ 5 h early

Kimi K3, and what we can still learn from the pelican benchmark

3 experts across 2 network communities independently surfaced this.

3 experts 2 communities 1 sources clustered
Simon Willison Ramon Astudillo ironyuppie.com

“My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even while it becomes further detached from how good the models are at the things that matter (like agentic tool calling across longer conversations) simonwillison…”

“simonwillison.net/2026/Jul/16/... >The new model is notable for the pricing: $3/million input tokens and $15/million output tokens, putting it at the same level as Anthropic’s Claude Sonnet series and making it the most expensive model released by a Chinese…”

2 experts discussed this · 3 posts
Simon Willison: My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even while it becomes further detached from how good the models are at the things that matter (like age…
Ramon Astudillo: My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even while it becomes further detached from how good the models are at the things that matter (like age…
Ramon Astudillo: death of the pelican test >That connection has been mostly severed now. The GPT-5.6 and Claude Fable 5 pelicans are outclassed by GLM-5.2, and much as I love GLM I don’t think that’s a Fable-class …
Open the full discussion →
Primary source simonwillison.net ↗
Established Models & releases Release 7d ago
xAI Grok Build coding agent released

xai-org/grok-build, now open source

2 experts are actively discussing the implications.

1 expert 1 community 1 sources clustered
Simon Willison

“I poked around in the just open sourced Grok Build CLI tool - 844,000 lines of Rust code! - and dug up a few interesting highlights, including a "self-contained terminal renderer for Mermaid diagrams" that renders them using Unicode box-art! simonwillison.n…”

2 experts discussed this · 5 posts
Simon Willison: I poked around in the just open sourced Grok Build CLI tool - 844,000 lines of Rust code! - and dug up a few interesting highlights, including a "self-contained terminal renderer for Mermaid diagra…
Luis Villa: @simonwillison.net no mention one way or the other in your commentary: I assume not a reproducible build, so no way to know if that code matches what is actually running?
Simon Willison: I didn't see anything that would help verify if it matches the xAI binary
Open the full discussion →
Primary source simonwillison.net ↗