[ExI] spike turing test, was RE: ai class at stanford

[Kelly Anderson]

On Fri, Sep 2, 2011 at 9:37 AM, Mike Dougherty <msd001 at gmail.com> wrote:
> On Fri, Sep 2, 2011 at 11:15 AM, Kelly Anderson <kellycoinguy at gmail.com> wrote:
>> On Thu, Sep 1, 2011 at 10:45 PM, spike <spike66 at att.net> wrote:
>>> I suppose the computer guided planes use lookup tables derived from how
>>> humans have played during the development phase, but the point is that with
>>> a sufficiently large lookup table, a computer is indistinguishable from a
>>
>> Large lookup tables can be done, but are highly inefficient. I'm sure
>> they have a slightly more elegant solution than that. You could
>> probably codify a pretty good fighter pilot with less than 100 rules.
>>
>> 1) Try to get behind the other guy (at least for WWI and WWII era fighting)
>> 2) Try to stay out from in front of the other guy.
>> 3) Shoot where the other guy will be when your bullets get there
>
> Kelly, I had a similar reaction to spike's repeated use of "large
> lookup tables" as a generic solution to any problem and was about to
> comment when I saw you beat me to it.  Ironically then you enumerate a
> list of tactics as an alternative to giant lookup table - which is
> effectively a lookup table of a slightly-higher-level rules.  True
> it's not a tape with a 'giant lookup table' of states and actions to
> apply  (which I think is where spike's coming from)

Pretty much any computerized solution can be converted to a "large
lookup table" problem if the lookup table is sufficiently large.
Unfortunately, in many common cases, sufficiently large would mean
employing every atom in the observable universe to storing the bits
necessary to hold said table. As a quick example of how fast it can
get complicated, consider a 640 x 480 x 24 bit color image.. using
that image as a key to a lookup table requires a table with a little
over 5 trillion entries. And my math may be wrong, because that
doesn't seem like enough... I'm a little foggy tonight.

> Spike, I assume you're familiar with Conway's Game of Life (cellular
> automata).  I like the idea that complex behaviors can be discovered
> (or emerge) on top of very simple rules.  Another great example you
> may not be familiar with is "boids" (use your favorite search
> appliance to find an online sim)  The algorithm models the behavior of
> a bird flock - with very simple rules for each bird.  The chaotic
> appearance of flock behavior begins to closely approximate the
> characteristic of real bird flocks (which I find fascinating,
> especially with hundreds or thousands of birds)  In OOP parlance, you
> could probably subclass the boid with a few fighter rules and do a
> convincing model of your WWII game.

Yes, it was exactly this bird flocking behavior (that I was aware of
before) that led me to post what I did... :-)   Great minds do, it
seems, work alike...

-Kelly