🔜@I've been increasingly saying "fitting to data" instead of "learning" when talking about ML. It feels like a more accurate reflection of what's actually going on (and gives more credit to the physics that it was originally drawn from)
🛄 chitter.xyz@parataxis after all, we're using Darwinian fitness algorithms to simulate learning, not anything resembling actual learning, right? =3
@Efi ehh, genetic algorithms aren't used much anymore. this is "curve fitting" not "fitness"
@parataxis augh! now I feel old nwn"" <3