From: Stephen Reed (reed@cyc.com)
Date: Wed May 22 2002 - 14:39:23 MDT
On Wed, 22 May 2002, Christian L. wrote:
> >And even if a spectacular breakthrough occurs, it will quite likely not be
> >interpreted properly.
> >
> >The step from an almost-human-level AI to a self-modifying AI with
> >exponentially increasing intelligence rapidly becoming superhuman, is
> >probably a step that most people won't be willing to take even when they
> >see the almost-human-level Ai ...
>
> Yes, I can see the arguments right now: "If it took like 50 years to get to
> almost-human-level AI it must be at least another 200 years to get even
> close to superintelligence..."
I would say that the recent experience of Y2K shows how awareness can be
propagated by opinion makers and decision makers when scientific evidence
points to a future high-impact event.
Consider the impact of superhuman AGI, as you all have been doing here,
it is simply stunning -- changing everything. At the very least
accounting firms will have to prepare business managers for the tremendous
increase in productivity, and the great reduction in labor cost as a
percent of global output.
Imagine a situation, some time in the future in which near human level AGI
has been achieved and in the news. Suppose that respected scientists are
making consensus forecasts that super human AGI is going to happen within
five more years. Given our Y2K experience, every company, every
government, every institution will have to prepare. In the case of Y2K,
it turned out that those who did not prepare fared OK for the most
part. But for what we are talking about there is many orders of magnitude
more impact.
The public will wake up if there is enough time between near-human AGI and
what follows... in my opinion.
-Steve
--
===========================================================
Stephen L. Reed phone: 512.342.4036
Cycorp, Suite 100 fax: 512.342.4040
3721 Executive Center Drive email: reed@cyc.com
Austin, TX 78731 web: http://www.cyc.com
download OpenCyc at http://www.opencyc.org
===========================================================
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:39 MDT