From: Ben Goertzel (email@example.com)
Date: Wed May 21 2003 - 11:16:37 MDT
> Now, of course if you are familiar with the way research progresses, you
> know that the power of public research comes from the lack of bureaucratic
> regulation and the reliance on peer review, collaboration and public
> disclosure. This is why AI's from this program would be better than AI's
> built in little closeted groups, and since only peer-reviewed programs (as
> opposed to bureaucrat-reviewed ones) would be funded, there would be the
> best balance between safety and efficiency.
Rafal, you have some solid and valid points here, but you should also recall
that the track record of heavily government-funded AI research is not very
The AI projects that governments have poured huge amounts of money into,
have generally turned out pretty disappointingly, given the cost. (This is
not to say that no progress has been made -- of course there have been some
real discoveries based on heavy gov't funding.) And an awful lot of the big
innovations have come from out-of-the-mainstream researchers who were
minimally funded or unfunded.
The US military has been the source of the vast majority of government AI
funding, and they have displayed a fairly strong bias toward expert systems
and logic based AI systems, which in my view has been detrimental to the
progress of AI. This bias has existed over a period of decades, in spite
of relatively unimpressive results.
If it is true that the mainstream of the AI field holds wrong ideas about AI
(as I believe to be the case), then unfortunately government funding is
unlikely to lead AI into the grand and glorious future.
What I think will happen is this. A maverick group (unfunded or minorly
funded by the gov't) will make a big breakthrough that clearly gets us
halfway or more to real AGI. Then the government will wise up -- because
the mainstream of AI researchers and their representatives in gov't are not
stupid or poorly-intentioned, just sometimes inflexible and biased -- and
start funding the breakthrough approach and other approaches like it.
-- Ben G
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT