From: Paul Hughes (psiphius@yahoo.com)
Date: Wed Apr 07 2004 - 12:32:59 MDT
--- Michael Wilson <mwdestinystar@yahoo.co.uk> wrote:
> Neither document is a constructive account, as I
> understand it for
> three reasons; risk, time and expertise. The risk
> factor is the most
> important; handing out detailed blueprints for
> building a seed AI is
> dangerous even if they're seriously flawed.
Why don't we take a different tact? I suggest, and
have already informed Eli and Tyler, that a good way
to approach this going forward, would be to place a
high priority on *friendliness* design and public
relations.
At the moment, the Singularity movement is ripe for
some serious misunderstanding in the public’s eye.
Already most Transhumanists have some serious
reservations about the whole seed-AI concept. Imagine
how magnified such misunderstanding will occur as
these ideas begin to percolate outside of our small
community.
So far the main theme of the Singularity movement has
been towards creating a seed-AI with friendliness.
Although ultimately creating the right kind of seed-AI
is our aim, wouldn't it behoove us to start a more
practical lets-built-it-now approach to engineering
friendliness into our intelligent systems as soon as
possible?
What I'm talking about is a shift, a moderate change
in perspective, in theme and perspective, towards
friendliness being our main goal (for now).
Starting with Friendly UI, moving on to:
Friendly Avatar
Friendly Believable Intelligent Agents
Friendly Intelligent Agents
Friendly Human-like Intelligent Agents
Friendly Above-Human Like AI
and finally,
Friendly Seed-AI.
I think such a shift would also shift the dialog away
from the seed AI, and most everyone’s fears of a
premature seed-AI launch, and towards focusing on how
to start engineering friendliness into everything NOW.
If we make friendliness our priority, then the
argument will be contained in the HOW, not the why, or
when, or if.
Lets built friendliness into our increasingly
intelligent systems now, and lets start discussing how
to do this. There are thousands of researchers working
on making AI smarter, but not necessarily more
friendly. This should be our call to arms, this
should be our mission, and this should be our goal.
Meanwhile, keep working on making the systems more
intelligent, but lets make friendliness are mission.
Paul Hughes
http://futurehi.net/
__________________________________
Do you Yahoo!?
Yahoo! Small Business $15K Web Design Giveaway
http://promotions.yahoo.com/design_giveaway/
This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:22:35 MST