[paulrosenzweig]

> "Deep Learning", no matter how big your dataset, fundamentally cannot possess this capability.

Why not? I’d agree that current models can’t, but what’s the fundamental shortcoming of deep learning?

[ibeckermayer]

The fundamental shortcoming is that there can be no understanding, and therefore no judgement, because there is no mind. See Searle's [in]famous Chinese Room Argument (https://plato.stanford.edu/entries/chinese-room/). Or for a more concrete example from today: https://twitter.com/mit_csail/status/1121447605412741120?s=2...