4 ms·

You’re absolutely right and I’m sure that something resembling higher-level pattern matching is present in the architecture and weights of the model, I’m just s

You’re absolutely right and I’m sure that something resembling higher-level pattern matching is present in the architecture and weights of the model, I’m just saying that I’m not aware of “logical thought” being explicitly optimized or designed for - it’s more of a sometimes-emergent feature of a machine that tries to approximate the content of the internet, which for some topics is dominated by mostly logical thought. I’m also unaware of a ground truth against which “correct facts” could even be trained for..

Closi 3y ago
> I’m also unaware of a ground truth against which “correct facts” could even be trained for.. Seems like there are quite a few obvious possibilities here off the top of my head. Ground truth for correct facts could be: 1) Wikidata 2) Mathematical ground truth (can be both generated and results validated automatically) including physics 3) Programming ground truth (can be validated by running the code and defining inputs/outputs) 4) Chess 5) Human labelled images and video 6) Map data 7) Dependent on your viewpoint, peer reviewed journals, as long as cited with sources.