RE: [SeattleRobotics] Humanoids
> -----Original Message-----I am quite willing to take your bet. How much money should we plunk down
> From: Bryan Andersen [mailto:bryan@...]
> I'm betting we will. My best guess looking at current trends in
> computer power and AI research that within the next 20-30 years there
> will be a few fully autonomous robots that if you interacted
> with them
> you would be very hard pressed to tell they are artificial.
> They will
> be to the point where they can be given a general asignment
> set and will
> follow it for months untill they need service. They will be fully
> capible of taking care of all their daily needs on their own
> and be able
> to request service on their own without human intervention.
Current trends in computer power and AI research are mostly worthless as an
indicator to the future. I have been listening to this same prediction for
at least 25 years. Today, we are not really any closer to the stated goal
than we were 25 years ago. We have basically moved about a centimeter in the
kilometer long race.
The computer power argument is the most worthless part. If someone could
demonstrate a computer program that was able to understand its surroundings
on a human level, have the ability to learn concepts on its own, and form
its own opinions and express creativity at a human level, I would be willing
to wait an hour or a day for the result. To date, nothing comes even close.
You could square the current computer power, and we still would be no closer
to a solution. You would just arrive at a non-answer faster!
This ain't a hardware problem! The hard parts of robotics are not mechanics
or control hardware, it's the control and intelligence software that we
don't know how to write. We don't know how to model human intelligence or
creativity. We certainly don't know how to write such a program.
I watch these television programs with great interest. However, don't get
yourselve too wrapped up in the hype. Asimo, Aibo, Cog, and all the rest of
the fantasic robots on the television are interesting achievements, but
their value is highly overstated.
I would love to be proven wrong.
- Five years ago, Most Everyone was thinking in serial. Your living in the
>To: "firstname.lastname@example.org" <email@example.com>
>Subject: Re: [SeattleRobotics] Humanoids
>Date: Thu, 05 Jun 2003 12:42:03 +1200 (NZST)
>I have been intrigued by the debate between the "hardware is not up to the
>and the "software is not up to the task" camps.
>What about this for an idea, perhaps it is the nature of both the current
>hardware and software that is the problem.
>Today pretty much every thing happens in serial. Hardware is serial,
>think in a serial fashion etc... We have psedo parallel execution provided
>operating systems, and multiprocessor machines, but it doesn't even begin
>approach the parallel processing power of the brain.
>At least, that is my uneducated understanding ;-)
>Visit the SRS Website at http://www.seattlerobotics.org
>To unsubscribe from this group, send an email to:
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
MSN 8 with e-mail virus protection service: 2 months FREE*