Comment by banachtarski

ML requires a useful corpus of data to train on. I don’t see how the energy abuse of crypto mining is remotely replicable with ML. For me, this has nothing to do with having a “neo-Luddite stance”