|
|
|
|
|
by visarga
630 days ago
|
|
> fMRIs say where blood is flowing in the brain, but that's hardly enough to explain the phenomenon of subjective experience. fMRI scans correlate well with neural net embeddings. That is a great hint. We just need to look at the semantic spaces developed in these models, by a purely mechanistic process, to see how it goes from data to semantics. |
|