[ExI] Automated black-box-based system design of unsupervised hyperintelligent learning systems

[Mike Dougherty]

2011/9/19 Amara D. Angelica <amara at kurzweilai.net>:
> Does anyone know of a system for automated black-box-based system design of
> unsupervised intelligent hyperintelligent learning systems?
[snip]
> All ideas, especially skeptical ones (so I don't waste time on blind alleys)
> would be greatly appreciated.
>
> Yes, this is a totally insane idea, but is it insane enough?

You can't build something smarter than yourself.  You might be able to
evolve something that gets smarter.  It'd be a good idea to start on
that as soon as possible.  All things being equal, anyone who has
started sooner than you has already won.  Your only advantage would be
if you are sufficiently smarter (clever-er) than they were at start.

If you're preparing to examine aliens, you might want to use
crowdsource computing and instead advance the world's communication
network(s) so humanity is better able to exchange information in real
time.  Until your supercomputer has enough on-board computing power to
outmatch humanity and its data collection/processing ability (7billion
nodes, each capable of observation/synthesis of ideas, etc) then we
might well use people for part of the solution.

Insane enough?  no.  add more insanity; I can still understand what
you're talking about.