All these AI companies MASSIVELY pirated the internet for training data
I think it’s worth being a little more careful with our phrasing here. AI companies definitely did illegal piracy to train their models, but that is a separate and distinct process from them trawling the web en mass. Anthropic, for example, recently settled a case with book publishers because it had pirated their books illegally. (Frustratingly, the second article that popped up when I searched to verify this was one that highlighted the fact that the TERF-in-chief is among the beneficiaries of that settlement.) But there is no clear legal answer to the question of whether crawling websites and training from their data constitutes even a civil tort, let alone illegal behaviour.
Other than that detail though, I think we’re in furious agreement. You’re simply restating my point, Cory’s point, and the subject of this article’s point, while doing so in a tone that implies disagreement. The problem isn’t the tech, it’s the humans using and creating the tech and the sociopolitical environment in which it happened.
I think it’s worth being a little more careful with our phrasing here. AI companies definitely did illegal piracy to train their models, but that is a separate and distinct process from them trawling the web en mass. Anthropic, for example, recently settled a case with book publishers because it had pirated their books illegally. (Frustratingly, the second article that popped up when I searched to verify this was one that highlighted the fact that the TERF-in-chief is among the beneficiaries of that settlement.) But there is no clear legal answer to the question of whether crawling websites and training from their data constitutes even a civil tort, let alone illegal behaviour.
Other than that detail though, I think we’re in furious agreement. You’re simply restating my point, Cory’s point, and the subject of this article’s point, while doing so in a tone that implies disagreement. The problem isn’t the tech, it’s the humans using and creating the tech and the sociopolitical environment in which it happened.