Training ever bigger convnets and LSTMs on ever bigger datasets gets us closer to Strong AI -- in the same sense that building taller towers gets us closer to the moon.
Show this thread
Remember that in the 70s and 80s we were supposed to reach strong AI by scaling up expert systems. Isn't human intelligence just an expert system with a really big ruleset and a really large knowledge base?
(
== 
)
-
-
jeopardy watson is arguably an example of this
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Just as we can make better planes by adding engines...pic.twitter.com/lykMvMcxGq
-
Excuse me, I am an expert in this aerodynamix thingy thing and can confidently tell you with enough money for big propellers I can get you to the moon. Two engines get us to 30k feet _easy_ NASA says the moon is only 1.261 billion feet away .: we only need 42k engines QED ^_^
- Show replies
New conversation -
-
-
Yes, and once a problem becomes too hard, you can distribute calculations across multiple people. That's why scientific discoveries 500 years ago were being made by individuals, while today you need a team to move even an inch in the right direction.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Isn’t “strong ai” just linear, brute force search but with a super fast processor?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
+ Imagination
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Taking it back, it would be exhausting to evaluate all the rules on all the knowledge for all the problems. We have many shortcuts to conserve energy for short term survival.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
But the difference I think is we can ourselves form new rules and decide the order of importance of the information we manage. Machines are starting to get there though I feel with modern advancements in DeepRL.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Any suggestions to us readers what should we be focusing on? We know picking low hanging fruits is low and hated upon. But we genuinely don't know which are the high hanging ones...
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.