Remix.run Logo
▲ dada216 an hour ago

In this specific case there is quite a simple solution, google (and oai/anthropic as well) should share revenue with the website they scrape to provide the ai summaries, they literally did this with adsense and youtube!

Sure, there's some technical hurdles, like identifying who's actually using the info they read, but impressions are a thing in ad networks, then there's broader questions about evaluating what training data was useful and used to ultimately produce what token, I'm still just taking about the kind of ai summaries in search related to the article, seems doable with the same system we use for ad networks.

When it comes to coding and more generalized knowledge there's already been interpretations about copyright laws and fair use that seem to allow what oai and ant are doing and how they are capitalizing on it, doesn't mean we should rethink them.

I'm sure there are technical solutions that can be translated to this sort of issues as well that could be derived from ad networks, I mean, imagine a world where your code and your blog post get a score on their usefulness to train LLMs and you get a share of the revenue, wouldn't this be better for all players? they get good training data, we get paid for it.

▲cromka 16 minutes ago | parent [-]

> should share revenue with the website they scrape to provide the ai summaries, they literally did this with adsense and youtube!

https://x402.org can facilitate that, we just need forcing the Big AI to pay out.