2 ms·
I think this will be the new normal. There are a lot more AI projects hungry for data to train their models on. This puts content companies in an uncomfortabl
by raid2000 3y ago
I think this will be the new normal.
There are a lot more AI projects hungry for data to train their models on. This puts content companies in an uncomfortable situation: trademark infringement claims, loss of intellectual property, and more.
- ffgjgf1 3y agoBut Github is not a content company and they don’t really own copyright to almost anything hosted there.
- raid2000 3y agoThat's true, but there's an interesting parallel with GitHub's corporate parent, Microsoft, and Microsoft's other platform company LinkedIn[1]. LinkedIn sued scrapers for retrieving data from the site. LinkedIn isn't a content company either, nor do they really own any content posted there (they don't right?), but a large part of their business moat comes from the network of people posting content there. Scrapers and bots undermine this, something the AI boom facilitates. 1: https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn
- al_borland 3y agoThere is a cost to serving up all that content, and if hundreds of AI start ups are all trying to pull in data, that can add up fast. It’s not typical user behavior.
- ffgjgf1 3y agoIf it’s just static content it wouldn’t really be that expensive. In reality egress traffic is extremely cheap compared to what Azure/AWS etc. are charging.