Is this safe/prudent ? (was Re: Perl AI Weblog)

From: Brett Paatsch (bpaatsch@bigpond.net.au)
Date: Mon Aug 11 2003 - 21:46:21 MDT

  • Next message: Eliezer S. Yudkowsky: "Re: Is this safe/prudent ? (was Re: Perl AI Weblog)"

    A.T Murray writes:

    > Artificial General Intelligence (AGI) requires extreme
    > programming efforts, whether in Perl or in other XYZ
    > programming environments.
    >
    > http://mind.sourceforge.net/perl.html is the new Perl
    > AI Weblog.
    >
    > No matter which AI language you favor, please accept
    > the challenge of coding a main Alife Mind program loop
    > with stubbed-in calls to the following artificial intelligence
    > mind-modules:
    > - Security
    > - Sensorium
    > - Emotion
    > - Think
    > - Volition
    > - Motorium.
    >
    > Please code this first AI module in your favorite
    > XYZ programming language by following the steps at
    > http://mind.sourceforge.net/acm.html -- DIY AI.
    >
    > If you know several programming languages,
    > please release the AI Mind loop in each of them.
    >
    > Put the resulting "main Alife Mind program loop"
    > on your personal Web site as open source AI code.
    >
    > Then come back to this forum and post a link to your
    > free AI source code for others to copy and enhance.
    >
    > The idea is, if enough AI code warriors implement the first
    > Mind-module, other AI enthusiasts will respond to the grand
    > challenge of AI and code the AI Mind a little further.
    >
    > We do not wish to standardize or homogenize the resulting AI code.
    > On the contrary, we want to see many pathways of AI evolution
    > branching off into the future in a survival-of-the-fittest race
    > towards the Technological Singularity of Vernor Vinge.
    >
    > A.T. Murray

    If I was a socially dysfunctional programming wiz with a burning
    desire to make my own avenging bot to settle some scores
    would not this sort of open source open sharing empower me
    in dangerous ways.

    I don't know if "friendliness" can be built into an AI, but
    I don't doubt some folks with knowledge will use IT savvy
    for mischief.

    Any thoughts on this sort of approach from a public
    policy stance Eliezer? Anyone?

    Regards,
    Brett

    This archive was generated by hypermail 2.1.5 : Mon Aug 11 2003 - 21:52:44 MDT