*SHOULDN'T BE CONTROVERSIAL* Many symbolic reasoning proponents argue for innateness. This myopic view misses the real magic of human intelligence: something that neither GOFAI or DL tackles: the ability to *discover* the categories worth reasoning about in the first place.

Put aside the innate climbing prowess of mountain goats. No human was ever born with the concept of *oxygen* or *the atom*. A truly ambitious AI agenda might benefit from looking ***less at beasts*** and ***more towards philosophy of science***.

We should ask where symbols and relations come from, how do they inform our exploration of our environments, and how do our observations, in turn, inform revisions to our systems for describing the world.