I think we can identify 3 levels of learning: 1 direct: input/output/feedback; 2 predictive: the input/output/feedback is internalized (modeled) so choices/actions can react to futures before they happen; 3 imagine: the models can be disconnected from actual input = intelligence.
-
-
-
Lately, we have seen some level 2 systems. Though they are not entirely neural net based, often the feedback is still applied from the outside (still called unsupervised learning). AlphaGo Zero is an example: https://en.wikipedia.org/wiki/AlphaGo_Zero …
End of conversation
New conversation -
-
-
The holy grail of artificial neural nets is how multiple nets should be connected together as in the brain. I think that's what they are getting at. I've been trying for 15 years as a hobby with no real success.
- 1 more reply
New conversation -
-
-
Will A.I machines be able to work together in a spontaneous way? - that is the question. Mankind has only achieved what it has by inspired collaborations, accidental meldings and competitive natures.
- 1 more reply
New conversation -
-
-
They might beat Kasparov at chess, but would be unable to have the vision to, say, demand a new chair for the match against Spasky.
-
Does emotional sacrifice equate the win? Is winning intelligence?
- End of conversation
New conversation -
-
-
Are there any developments in cognitive science that have contributed as much to our understanding of intelligence as ANNs?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
As machines get smarter, humans are getting dumber. Cause n effect?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Ten more to go.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
And I predict in far less than 30 years, we all will think very differently (AIs and humans).
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
True. ANNs supply intelligence that humans don't have. Solving general AI will require the opposite. Will happen though.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Tweet mark
-
1) Extrapolating from incomplete data is... the only way to function in this world we live in. I've figured that we (humans) have used our imagination to fill in the gaps. Good to create stories to piece things story, bad for our potential for delusion. A broader definition.
-
2) The definition of delusion Buddhists use... which is anything in our imagination that doesn't match the natural world. We are all delusional... especially in the social sciences. The hope is that AI can have better methods of reality testing built into their imagination.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.