Substack: dataleverage.substack.com/p/how-collec...

Leaflet: dataleverage.leaflet.pub/3m2wpj7l7c22w

Also just saw this paper this week and was similarly quite excited and thinking along similar lines!

Follow up, tying together "AI as ranking chunks of human records" with "eval leverage" and "dataset details as quality signals": dataleverage.substack.com/p/how-do-we-...

And related, "eval leverage": dataleverage.substack.com/p/evaluation...

(1) ongoing challenges in benchmarking, (2) challenges in communicating benchmarks to the public, (3) dataset documentation, and (4) post-hoc dataset "reverse engineering"

The original post: dataleverage.substack.com/p/selling-ag...

who paid that Dr for a verified attestation with provenance can use this attestation as a quality signal; a promise to consumers about the exact nature of the evaluation. A "9/10 dentists recommend" for a chatbot.

More generally, I think there are interesting connections between current discourse &

For some types of info, we can maybe treat as open and focus on selling convenient/"nice" packages (ala Wikimedia Enterprise)

But attestations provide another object to transact over. Valuable info (a Dr giving thumbs up/down on medical responses) may leak, but the AI developer

So in a post-AI world, to help people transact over work that produces information, we likely need:
- individual property-ish rights over info (not a great way to go, IMO)
- rights that enable collective bargaining (good!)
- or...

The core challenge: many inputs into AI are information, and thus hard to design efficient markets for. Info is hard to exclude (pre-training data remains very hard to exclude, but even post-training data may be hard without sufficient effort)

It looks like some skepticism was warranted (not much progress towards this vision yet). I do think "dataset details as quality signals" is still possible though, and could play a key role in addressing looming information economics challenges.

Finally, I recently shared a preprint that relates deeply to the above ideas, on Collective Bargaining for Information: arxiv.org/abs/2506.10272, and have a blog post on this as well: dataleverage.substack.com/p/on-ai-driv...

And we have a blog post on algorithmic collective action with multiple collectives! dataleverage.substack.com/p/algorithmi...

These blog posts expand on attentional agency:
- genAI as ranking chunks of info: dataleverage.substack.com/p/google-and...
- utility of AI stems from people: dataleverage.substack.com/p/each-insta...
- connection to evals: dataleverage.substack.com/p/how-do-we-...