Shane Legg wrote:
On 3/19/07, *Ben Goertzel* <[EMAIL PROTECTED]
<mailto:[EMAIL PROTECTED]>> wrote:
conservative incremental steps, the current scientific community
is highly culturally biased against anyone who wants to make
a large leap. Science has drifted into a cultural configuration that
is obsessed with making incremental progress with a very small
increment size.
I don't think it's because science is against large leaps, the large leaps
are what everybody in science loves.
The science establishment loves large leaps in hindsight ...
but yet in
foresight, follows a funding pattern (and publication-vetting) that is
oriented
almost entirely toward small incremental steps, instead.
There are exceptions of course, e.g. Human Genome Project
They (or perhaps I should say we)
want evidence, generally in the form of a proof or experimental
results that
can be repeated. Furthermore, the larger the leap the more impressive the
evidence has to be. Thus, if somebody says that they can build a thinking
machine with general intelligence equal to that of a human but don't have
amazingly strong evidence, nobody much will pay attention.
On the other hand if someone can demonstrate a working system with human
level AGI, they will have no trouble in getting scientific attention
and respect.
I have a couple responses to this:
1) As I said, large leaps are admired and celebrated in hindsight; but the
pattern of funding and publication-vetting is not at all designed so as to
encourage them prospectively.
2) The need for evidence and substantiation is interpreted in a highly
subjective
way based on prevailing theoretical paradigms. For instance, the Human
Genome
Project was funded with no hard evidence that it would be useful.
Instead, the leading
scientists sorta fooled the politicians into thinking tremendous
applications
would follow as soon as the sequencing was done. And now, not too
surprisingly,
it is taking loads more funding and time to get much real use out of it
(because,
as many foresaw, just knowing the gene sequence doesn't tell you that
much...
it's only a start...). And of course, string theory is well funded
within the physics
establishment now, in spite of zero empirical evidence and fairly weak
theoretical evidence. Other physics approaches with equal or greater
evidence
are not favored at the moment (ask Juergen ;-)
But the early days of string theory illustrate Point 1 above.
It's often been noted that string theory was originally
developed mainly by men in their 40's. This is because, given the culture
of the physics establishment and the difficulty of the physics job market,
it was too risky for young pre-tenure faculty to spend their time working
on something so "eccentric".... Now however string theory has become
mainstream (the initial large leap was already made, though in this case
it did not lead to any empirical verification, it was purely a conceptual/
mathematical leap), so that young profs can get away with working on it
without killing their careers.
Similarly, right now, AGI is a somewhat risk career move for AI profs
at the pre-tenure stage.
-- Ben
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?list_id=11983
