It doesn’t contradict the genomic encoding at all. But it puts a different emphasis as far as the need to *additionally* have a quite powerful RL system operating, to get the right behavior. It suggests a different AI emphasis: meta learning or evolving very specific reward fxns.
-
-
Replying to @AdamMarblestone @TonyZador and
Psychologists call this stuff "motivation"; it' part of the answer, but inot the whole answer. What I don't understand is why some folks are comfortable ascribing fairly detailed innate structure to motivations (or loss functions) but not to other aspects of cognition.
3 replies 0 retweets 2 likes -
Replying to @GaryMarcus @TonyZador and
I’m comfortable ascribing innate structure to a ton else. But loss functions are nice and compact, and we know there is learning... so what I’m less comfortable with is AI people using too-generic and math-y end to end losses rather than diverse ethologically specific losses.
1 reply 0 retweets 4 likes -
Replying to @AdamMarblestone @GaryMarcus and
The issue is that if you assume specific losses, which only act in specific settings, things start to just look like programs (in the computer sense.) Not a problem unless you want to eliminate explicit ops over variables, binding, etc. from cognition: but some seem to want that.
1 reply 0 retweets 1 like -
Replying to @recursus @GaryMarcus and
It looks like a developmental biology program! Also I should say that I do understand why people study simple end to end loss functions: to isolate other aspects of the machine learning problem and to keep things simple. So I don’t really mean that as a criticism of ML.
1 reply 0 retweets 0 likes -
Replying to @AdamMarblestone @recursus and
And yes I know
@GaryMarcus said it was a developmental biology program in one of his books, which I think is even more under emphasized than variable binding now.2 replies 1 retweet 3 likes -
Replying to @AdamMarblestone @GaryMarcus and
To take Hopi's lab's work as ex: len(burrow) could be outcome of a rewarding process or a "while loop." But in either case, need to specify why burrowing only kicks in at certain places, times, types of dirt, etc.; this seems more naturally a program than a set of loss fns.
2 replies 0 retweets 1 like -
Replying to @recursus @AdamMarblestone and
The “program” argument has been used since the beginning of ethology to describe species typical behaviors, just look at all the debate surrounding imprinting but how these debates largely glossed over developmental processes, Gottlieb and Bateson largely settled these arguments.
1 reply 0 retweets 1 like -
Replying to @DialecticalA @recursus and
I don’t understand why there would be any tension between innate behaviors and developmental programs. The one arises from the other. It’d be like arguing about how fractals can look different without talking about recursion functions and their parameters
1 reply 0 retweets 6 likes -
Replying to @TonyZador @recursus and
Because Innate and program have assumed some latent information exists and earlier developmental stage that simply unfolds to later stages, but that’s not how it works, development is interactions across multiple levels. I would highly recommend Mark Blumberg’s perspective here.pic.twitter.com/R6afqVMS0N
2 replies 1 retweet 4 likes
I doubt there is any real disagreement here. But emphasizing a rich “programming paradigm” for constructing “innate” behavior w/ programmed loss fxns, external data inputs, & staging thereof based on various forms of recognition, in addition to pattern generators, is relevant...
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.