[ExI] The present and future of AI

[John Clark]

On Fri, Jun 12, 2020 at 2:50 PM Jason Resch via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

*> I am curious how close members on this list feel we are from
> achieving AGI. Most estimates seem to put it around 2040 plus or minus 10
> years. Is that the general consensus? *


That would be my best guess but it's only a guess; the inherent iterative
nature of the thing makes predicting a future timeline difacult.

*> Could it happen much sooner? *


Sure, it could come suddenly out of the blue with little warning; if for
example somebody developed a computer program that could improve software
as well as Alphazero can improve gameplay for GO or Chess with no
assistance from humans needed. If that happened then we could wake up one
morning and find ourselves in the possession of a high level computer
language that was so easy to use that grade school kids could learn it with
no problem but ran as efficiently as the very best machine code produced by
expert human programmers could with the help of assembly language.

*> Does anyone here believe it will happen much later than that?*


It could be later but if it does I think it would be due to societal
factors and not because of technological limitations. I say this because
thanks to biology we know of an upper limit on the amount of complexity a
seed AI would need. In the entire human genome there are only 3 billion
base pairs, there are 4 bases so each base can represent 2 bits, there are
8 bits per byte so that comes out to 750 meg which you could fit on an old
fashioned CD audio disk. And there is a huge amount of redundancy in the
genome so with file compression you'd probably need less than a third of
that CD audio disk, and of course much of the genome is about stuff that
has nothing to do with the brain or intelligence.

John K Clark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20200612/14f838d9/attachment.htm>