[ExI] i got friends in low places

[spike at rainier66.com]

-----Original Message-----
From: extropy-chat <extropy-chat-bounces at lists.extropy.org> On Behalf Of BillK via extropy-chat
>
>>... We might as well just leave the AI goalpost on the back of the truck just to save time.
>
> spike
> _______________________________________________



<https://bengoertzel.substack.com/p/three-viable-paths-to-true-agi>
Quote:
>>... — but which, due to the fundamental lack of ability to innovate, abstract or generalize, would be incapable to address difficult unsolved science and engineering problems, or to perform the self-modification and self-improvement needed to serve as seed AIs and launch a Singularity... Ben
-------------

>...ML programs are very useful algorithms but they are not AI.

BillK
____________________________________________


BillK, the Kalman filter software does a kinda-sorta version of both self-modification and self-improvement.  It modifies itself by throwing out irrelevant correlations (butterflies) and adds in new observable metrics into the covariance matrix.  It self-improves by calculating, estimating, observing, adding the measured observations to data, recalculating correlation coefficients then replacing old coefficients with new ones in the covariance matrix.

Me lads, that is both self-modification and self-improvement.  Clarification: it is more self-modification and self-improvement than I am doing, for time is modifying me (in a way I don't like) and my self-improvement is negative because of it.  I didn't usta suck.  But now I do.

But I digress (again.)  If we get a jillion independent processors working on Kalman filters, each doing their own thing (hurricane predictions, my own favorite the predictions of the time interval in which the next record prime number will be discovered, predictions of droughts for three examples) then load up the accumulated expertise in one computer, and if we do that... what is that computer?  

Fun answer: it isn't intelligent because it isn't self-modifying or learning at all.  It just contains all the accumulated wisdom the Kalman filter computers already calculated and learned, thru their self-improvement and self-modification.

BillK, consider your comment about ML programs being useful algorithms but not AI.  You offered us no backup reasoning, which is what I am asking for right now.  Drill deep sir.  You might be right on that contention (for it is in agreement with some powerful thinkers (such as Eleizer and Ben)) but do support your conclusion.

spike