Simon Tesla 🏳️‍🌈🐭 @SimonTesla@chitter.xyz

So, uh, #introductions, I guess. I'm a #furry computer mouse :3

You ever think about how ironic the term "lingua franca" is? Especially as it's commonly used in the English language?

musings prompted by this article, originally found via Metafilter: https://waxy.org/2022/09/ai-data-laundering-how-academic-and-nonprofit-researchers-shield-tech-companies-from-accountability/

At the very least, humans are able to draw on experiences that ML is likely incapable of training on, and have the ability to bring their own judgement and cultural awareness to the table. To me this difference feels fundamental, rather than incidental.

I think one of the most disingenuous things about saying that it's ok to train AI using copyrighted images is that "it's exactly like what humans do to create art". The problem with this argument is that it's not falsifiable, at least not with a lot better understanding of how both systems work. Both ML and brains are still mostly black boxes at this point, and saying one works like the other feels obviously wrong, but not in a way that is easy to argue.