- @SimonXIX the value of humanlike ai is to better understand humanity. And that's all. And yes, every ai researcher knows this. in reply to SimonXIX #
- @DimSum111 name a non-humanistic construct. in reply to DimSum111 #
- @SimonXIX and have a whole set of new ones. And it takes us many years to make sense of the world, after eons of evolution. Hard. in reply to SimonXIX #
- @ollie_cheshire yes they can, built one myself that tells me no if it can't do it. in reply to ollie_cheshire #
- @SimonXIX if the system can operate in the world, it understands it. There's no other meaningful definition of 'understand'. in reply to SimonXIX #
- @mrscientist selection bias. Much more work is being done at animal level, but its not sexy enough for telly. in reply to mrscientist #
- @SimonXIX it's also the answer to the Chinese Room argument. in reply to SimonXIX #
- @ollie_cheshire not explicitly; it learned that it some cases it was a good idea to turn down a task rather than attempt it. in reply to ollie_cheshire #
- @Gion_Gion nothing us impossible. Pointless, yes. in reply to Gion_Gion #
- @Lissy_m true. But they do exist. My robotics lecturer, Myra Wilson, for example. in reply to Lissy_m #
Powered by Twitter Tools
