Re: Is a theory of hard take off possible? Re: Investing in FAI research: now vs. later

From: William Pearson (wil.pearson@gmail.com)
Date: Thu Feb 21 2008 - 05:31:46 MST


On 21/02/2008, Eliezer S. Yudkowsky <sentience@pobox.com> wrote:
> William Pearson wrote:
> >
> > Now imagine you added a link to the outside, through which one bit
> > could enter. Now depending upon that bit the system, the systems
> > evolution could bifurcate, or it could ignore it and stay singular.
> > Bifurcation leads to the potential for growing of the ability to
> > predict the world, if parts of the world happen to correlate with the
> > bit that was bifurcated on.
> >
> > More bits given to the system lead to more possible bifurcations.
> > Exponentially increasing numbers of bifurcations are needed for
> > exponential increases in predictive power.
>
>
> You need a logarithmic number of bits in order to slice a prediction
> to great fineness - i.e., predicting to within a factor of a billionth
> requires 30 bits, not a billion bits. I know this is probably what
> you meant, but nobody would hear you saying it unless they already
> knew the answer.
>
> The other thing to remember is that we do not live in a random
> universe where it takes 1 bit of information to predict one more bit
> of an important outcome, like trying to record a sequence of
> coinflips.

Humans sometimes try and make it so. See one time pads etc. Also if
you consider the ease that systems with feedback loops make UTMs you
need considerable information about the start states to be able to
predict non-trivial properties of it. Also in general to be able to
maintain predictive power over another bifurcating system, you have to
bifurcate as fast as it is bifurcating. How quickly is society
bifurcating? Is it exponentially increasing its rate of bifurcation?

> Gravity is the same all over, you don't have to learn it
> again each time; and many important and manipulable aspects of the
> universe are highly regular.
>

But you still need to get lots of bits to determine the gravity
problem you are working on.

>
>
> You are going down a correct path, mathematically speaking; but the
> answer is going to end up being "In principle, it takes a ridiculously
> small amount of information; the exact amount is theoretically
> incomputable; and a superintelligence would require more, but we can't
> guess how much more without actually running a superintelligence."

The answer to what question exactly? Also I am not comfortable with
superintelligence as a word. Is it possible to taboo it?

> A cable modem connected to the Internet is certainly *vastly* more
> bandwidth than is theoretically needed to grok and take over the
> world, though.

Takeover, I would grant you. You could bifurcate the world from the
state of being not taken over to being taken over with one bit, if the
world happened to be at that tipping point. But grok? Surely that all
depends upon where you start from.

> Or a 1200 baud modem, for that matter. Hell, give me
> a millionfold subjective speedup and a few siderial days to think
> about it and *I'll* take over the world with a 1200 baud modem (not
> because I want it; just to demonstrate the point).

Why wouldn't you want to? If you can take it over in a meaningful
sense you can quash all non FAI development and put lots more
resources to the development of FAI.

Also you have had significant bifurcations already. Give baby eliezer
a millionfold subjective speedup and a 1200 baud modem (and no eyes or
ears), and I would like to seem him take over the world.

I suppose I am interested in the transition time between the situation
we are now, computers are very bad at predicting the world, to the
situation where a computer is very good at predicting the world and
therefore being able to manipulate it with some purpose.

> Remember that there are more reliable ways of making money than trying
> to win the lottery - a wise agent can steer their pressure to the more
> manipulable, more predictable, lower-entropy levers in the system.
>
> When bifurcation is a limited resource, you can arrange to use less of
> it - you don't *have* to let 50% of your selves emit alternate
> versions of every bit you want to output.
>

I don't see how you can use less bifurcations and still *improve*.
That seems to me to be like trying to do non-experimental science.

  Will Pearson



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT