Re: Steve Harris, "Return of the Krell Machine" (Singularity Article)

From: Dan Clemmensen (
Date: Sat Jan 19 2002 - 09:58:10 MST

Paul Hughes wrote:

> --- Ben Goertzel <> wrote:
>>Actually, I don't agree with that! I think that is
>>one possible path to
>>creating a artificial intelligence, but not the
>>only possible path; and I
>>suspect that it *won't* be the first path followed.
>>I suspect that AI will
>>move faster than brain science in the next decade,
>>even though it's lagged
>>behind in the past 2 decades.
> I find this extremely doubtful. I have several
> friends who have gotten their hard-earned Ph.D's in AI
> and every single one of them says that creating an AI
> through experimental software methods alone and not
> aided by a thorough knowledge of brain science will be
> doomed to failure. In addition they also all agree
> that unlimited computer speed alone with do nothing to
> advance the state of AI Research. One of my friends
> recently told me that the biggest flaw with several AI
> pathways is their overestimation of their scalability.
> He compares them to a person who, climbing a tree and
> seeing that they are moving ever upward, concluding
> that if they continue they will eventually reach the
> moon. Truer words have never been spoken.

This whole argument misses the point: AI may not be the
only path to SI. I feel that we'll get SI via human-computer
collaboration. We don't need to re-create the qualities of
the human brain in the computer. Instead, we need to figure
out how computers can help people to be more intelligent.
If the correct programs can gather and organize information
for evaluation by a human, the human should become a more effective
thinker. I'm more effective because of Google. If
I really trained myself to use it consistently, I would
be even more effective. If I used Google to help me learn
to create search engine software and then to build a better
search engine, I would be even more effective, especially if
I optimized the engine for software development assistance.
I might next choose to develop "groupware" oriented toward software
development, and invite other humans to join. This effort might
result in a team that can go from concept to implementation very
rapidly. We use then spin off cooperating teams to improve data
presentation, to optimize software and operating systems, to optimize
networking, to optimize scalability, and finally to work in AI.
By that time. the collaboration may already be an SI, without
and will therefore be much more effective in solving the remaining
AI problems.

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT