Computational models of cognition:Reverse-engineering common sense in the human mind and brain Pt 1

[@leonlysak4927]

Wow what an underrated talk

[@CharlesVanNoland]

@3:50 I don't personally believe being able to share models socially is a requisite for useful and valuable machine intelligence. Animals convey only their coarse observations, i.e. inbound threats or food, but can understand rather complex aspects of their surroundings that imply world modeling is being utilized. This preoccupation with "human level" intelligence seems to be tangled up and confused with what would actually be useful as a minimum viable product. There are many applications where high level perception and motor skills and even problem solving would be a tremendous value without the agent(s) articulating to anyone or anything the how or why it chooses what it acts upon. We don't absolutely need machines that act and behave just like humans do. Value can be had by a vast array of different forms of machine intelligence without many of the attributes of humans.

[@jamesreagan9063]

Seems to me that is the inherent frailty of most human beings!

[@johnniefujita]

0:09:01 The only conclusion that i get is that waymo's cars are sensitive and beautiful. lol

[@paulbizard3493]

Very interesting.