Tootfinder

Cory stumbles a lot here under even Michael's light questioning.
Cory seems to think that the machines are just producing fake stories about how to plan a hack into Hugging Face and that somehow that isn't the same as planning a hack into Hugging Face.
I mean there is a difference between those things: between finding a path in a search tree of words about plausible plans and working out a plan anew, from a deep understanding of the world.
The difference does possibly mean that they can't try hacks nobody else has ever tried in the history of all mankind, whereas an expert person might think of a new one.
And there's a philosophical difference: whether they are thinking or just producing a stream of what someone might think if they were thinking.
But in almost all actual piratical cases, writing a plausible story about how someone might make a plan does end up with an actual plan, even if it ain't an original one. And it can easily be a plan the operator wouldn't have thought of.
RE: Cory Doctrow on Downstream.
#ai #coryDoctorow #novaraMedia #link