Joshua Fox wrote:
Richard, thanks for the reference to Kuhn. I was aware of his "paradigm
shift" concepts, although I have not yet read his writings.
Pondering this question, I came across an example in a field which I
studied.
To me, the fact that it is in another field provides some distance which
highlights the AGI situation. I hope that the following is not too far
afield for this list, nor too obvious to those knowledgeable in the
history of science. Maybe it's even interesting!
In diachronic linguistics, most scholars try to reconstruct an
unattested parent language only a thousand years or so before the
earliest attestations of language families, as for example,
Proto-Indo-European or Proto-Semitic. However, some eminent professors,
such as Aron Dolgopolsky, claim to be able to reconstruct far earlier,
to a proto-language which is the ancestor several steps back from the
more commonly reconstructed proto-languages.
What do the respected professors in the field think about these rebels,
called Nostraticists? (My views are identical to the establishment; I
claim no fearless independent-mindedness.)
They think that
- The Nostraticists are NOT pseudo-scientists. Their techniques are the
same as mainstream scholars, but they push the evidence to more
speculative conclusions. They do work that is on the edge of weird, but
only on the edge.
- The Nostraticists' reconstructions cannot be called provably wrong;
rather, they do not have the evidence for well-founded conclusions. They
are triangulating too far using noisy, sparse data and uncertain lines
of reconstruction.
- The Nostraticists may well be proven right. Indeed, similar hypotheses
about very large language groups gained acceptance by the consensus.
- Future methodologies may help, but we have absolutely no idea what
those methodologies might be.
- The Nostraticists are NOT Young Turks. In fact, their school had a
small golden age decades ago -- and the leading Nostraticists were
young, then -- but Dolgopolsky and others of the school are now old,
retired, or deceased. There may be a few new scholars in the field, but
it didn't take off.
- A grad student who is turned on by the Nostratic approach can seek out
a professor and study with them. This is very rare in practice. Such a
student will NOT automatically be considered an academic leper, but may
be left at the margins of the academic consensus.
- The Nostraticists' ideas excite a sense of wonder, even in the most jaded.
- Requests for speculation on the Nostraticists' far-out ideas will not
be met with energetic visions of future possibilities, but with a
cautious, lukewarm, "we just don't know." A layperson might be
disappointed at what appears to be a total lack of imagination and
intellectual courage.
The comparison to AGI is by no means exact, but it does highlight the
scholarly conservativism and the avoidance of speculative -- even though
scientifically based and potentially very fruitful -- paths of inquiry
Joshua
It is an interesting case, for sure (I had heard about the general ideas
of the Nostraticists, and I admit that they strike a deliciously
romantic chord ...).
But the correspondence with AGI is probably not close enough to help us
much, I think. Mostly because their idea is static: as you pointed
out, without further evidence there is not much anyone could say about
it. In AGI, on the other hand, everything remains to be discovered.
For me, the real problem with AGI boils down to something pretty simple.
The nature of the problem is such that it demands a new methodology.
That methodology (in the small possible nutshell: treating it as a
science, rather than an engineering or mathematical discipline) is so
repugnant to the people who dominate the field today that they will
never allow anyone to go that route, so the only thing we can say to
them is "Go away: Do something else, please."
Clearly they would never do this voluntarily. So they are going to have
to be pushed. But they shouldn't really be too unhappy about it: the
present generation of AI folks (the Mathematicians) did exactly this to
the previous generation of AI folks (the Engineers) 20 years ago ;-).
There are lots of other bits, but that is the core problem.
Classic case of a paradigm shift, as far as I can tell.
Somebody just has to say to them: "All Your Projects Are Belong To Us"
and that will be the start of a paradigm war.
Hey, come to think of it: consider it said.
Richard Loosemore.
-----
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
