Maybe this is grasping at straws, but it’s fascinating how close evolutionary processes of different systems mirror each other. (1/n)
-
-
In computing most search algorithms (take Djykstra’s as an example) are basically “rapid guess and check.” Machine learning and AI are essentially “guess and check” done incredibly quickly at a huge scale with beefed up computer hardware (3/n)
Show this thread -
So here’s my unifying theory of life: Systems are so complex that no one can predict anything, so basically we should try as many things as we possibly can and discard the losers as quickly as we can, knowing that learning takes so long we’ll all probably die first (4/n)
Show this thread
End of conversation
New conversation -
-
-
THIS IS DANGEROUSLY CLOSE TO BECOMING A BAD ANALOGY QUIT WHILE YOU'RE AHEAD
(there are people applying evolutionary ideas to machine learning, but IMO the common gradient-descent approaches are not particularly analogous to evolution, they're more analogous to *breeding*) -
Oh we crossed over the realm of bad analogy minutes ago. There’s no going back now.
End of conversation
New conversation -
-
-
I think there are parallels in every complex system - the ability for individual notes to die out (species in an ecosystem) / go bankrupt (businesses in an economy) renders the larger system more resilient. (Might have stolen this from Antifragile).
-
With this line of thinking, governments should’ve let the banks die after the crash for long term benefit. But if that nuked the economy short term, they’d be voted out. So even if they though this way, it may have been too risky for preserving their power.
-
*thought
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.