Oh the classic problem of being inside the system and being outside the system. May be we can AI to tweak the rules slightly, learn from the new system and do transfer learning.
-
-
We got out once we discovered the nature of universal computation, so we found an a priori frame that allows to deal with our minds independently from the substrate.
1 reply 0 retweets 4 likes -
That’s abstract knowledge, which does not accurately describe the real universe.
2 replies 0 retweets 0 likes -
Replying to @BTCMoonquake @scadza and
Did you find a tool outside of rationality that allows you to accurately describe the real universe? If so, you might want to consider that you were infected by a mind virus?
3 replies 0 retweets 0 likes -
Replying to @Plinz @BTCMoonquake and
What is to say that NNs would exhibit the neural disorders that plagues humans may be at the trillionth layer perhaps?
1 reply 0 retweets 0 likes -
Replying to @scadza @BTCMoonquake and
Our nervous systems have evolution calibrated hyperparameters that speed up convergence, and self-calibrating hyperparameters that depend on sufficient functionality of neurotransmitter metabolism. If one of these is off, you may get bad hyperparameters, even in a shallow net.
2 replies 0 retweets 1 like -
Replying to @Plinz @BTCMoonquake and
But we have abstracted out remember ? We not playing on the same substrate anymore.
1 reply 0 retweets 0 likes -
Replying to @scadza @BTCMoonquake and
If your brain cannot approximate the execution of its search algorithm for function approximators well enough, the abstraction breaks down, because you are no longer a universal computer for our purposes.
2 replies 0 retweets 0 likes -
When in evolution did we switch to being universal computers? This must be an emergent function of network size then, no? Is this in accordance with Evolution Theory, given that being an UC seems such a strong selector and Dolphins et al already own the necessary hardware?
1 reply 0 retweets 0 likes -
Replying to @jasogamol @Plinz and
It is vitally important to get this - naturalistic neuroscience is at an impasse until and unless it accepts that hardware (neurons) do not generate mind. The “Hard Problem” will not be solved by reductionist positivism.
2 replies 0 retweets 0 likes
I believe that it is solved. Consciousness is (as we can experimentally show) not an experience of the "here and now" but the memory of an experience that never was. Functionally it is a model of the contents of our attention (including reflexive), used for credit assignment.
-
-
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.