New Framework Creates Markets For AI Pretraining Data

@derekslater @HarvardBiz Actually, solution to this is built in: funds are directed by the model makers, to maximize data value...

@glenweyl @HarvardBiz really interesting stuff. My worry is this still in practice ends up as just a wealth transfer to large media & CMOs, in part because most works are not actively managed / licensed - see e.g. https://www.promarket.org/2025/11/19/the-false-hope-of-content-licensing-at-internet-scale/ - but your point is much more narrower than that

@glenweyl @HarvardBiz Not following, maybe I'm missing something tho - how would the CMO get the rights to all the unmanaged works, and then distribute the money to people who haven't come forward?