[ExI] Watson on NOVA

Kelly Anderson kellycoinguy at gmail.com
Mon Feb 14 17:02:19 UTC 2011

On Mon, Feb 14, 2011 at 6:24 AM, Richard Loosemore <rpwl at lightlink.com> wrote:
> Kelly Anderson wrote:
> :-)
> Well, first off, don't get me wrong, because I say all this with a smile.
>  When I went to the AGI-09 conference, there was one guy there (Ed Porter)
> who had spent many hours getting mad at me online, and he was eager to find
> me in person.  He spent the first couple of days failing to locate me in a
> gathering of only 100 people, all of whom were wearing name badges, because
> he was looking for some kind of mad, sullen, angry grump.  The fact that I
> was not old, and was smiling, talking and laughing all the time meant that
> he didn't even bother to look at my name badge.  We got along just great for
> the rest of the conference.  ;-)

I'm glad to hear you aren't grumpy in person... but you do come off
that way online.. :-)

> Anyhow.
> Just keep in mind one thing.  I criticize projects like Watson because if
> you look deeply at the history of AI you will notice that it seems to be an
> unending series of cheap tricks, all touted to be the beginning of something
> great.   But so many of these so-called "advances" were then followed by a
> dead end.  After watching this process happen over and over again, you can
> start to recognize the symptoms of yet another one.

I understood this to be your position.

> The positive spin on Watson that you give, above, is way too optimistic.  It
> is not a parallel approach, valid and worth considering in its own right.
>  It will not make IBM any money (Big Blue didn't).  It has to run on a
> supercomputer.

Google runs on a supercomputer too. The same basic kind of
supercomputer. Also, an iPhone has the computational power of NORAD
circa 1965... so lots of extra computation can buy you a lot, even if
not AGI all by itself.

> It is not competition to any real AI project, because it
> just does a narrow-domain task in a way that does not generalize to more
> useful tasks.  It will probably not be useful, because it cheats:  it uses
> massive supercomputing power to crack a nut.

I think answering questions is a generally useful task.

> As a knowledge assistant that could help doctors with diagnosis:  fine, but
> it is not really pushing the state of the art at all.  There are already
> systems that do that, and the only difference between them and Watson
> is..... you cannot assign one supercomputer to each doctor on the planet!

Of course you can. Put it online, time share it, put it in the cloud.
All this works fine. Most doctors wouldn't use such a system for more
than a few minutes a week since most of their work is pretty routine.

> The list goes on and on.  But there is no point laboring it.
> Here is my favorite Watson mistake, reported by NPR this morning:
> Question:  "What do grasshoppers eat?"
> Notice that this question contains very few words, meaning that Watson's
> cluster-analysis algorithm has very little context to work with here: all it
> can do is find contexts in which the words "eat" and "grasshopper" are in
> close proximity.  So what answer did Watson give:
> "What is 'kosher'?"
> Sigh!   ;-)

As for IBM making money from Deep Blue, I would ask did Americans
benefit from the space program? Research isn't made to directly make
money, but to lead the company in directions that will make money.
Last time I checked, IBM was still profitable. Without research, they
soon would not be profitable.

What Watson tells the world is that IBM is still relevant. If that
supports their stock price, then the Watson team has earned their
money. There are now world class chess programs that run on cell
phones. In ten years, there will be Watson like programs running on
cell phone sized devices, but working better.

I'm not impressed by Watson mistakes. We KNOW it isn't intelligent, it
just does what it does better than most humans. Over the next three
days, we'll see if it does what it does better than the very best
humans. Ken Jennings lives around here somewhere. I am kind of
surprised I've never run into him.


More information about the extropy-chat mailing list