| 
AI Ramblings 
Thursday 06 November 2008  Almost right...
I found a paper by Diederik Aerts about composite concepts recognition which in spite of the mouthful of the title (General Quantum Modeling of Combining Concepts: A Quantum Field Model in Fock Space) hints at a very simple and always overlooked fact: We do not use any "obvious" logic in our actual appraisal of category membership.
Following several previous papers of his own about the use of quantum logic instead of classical (boolean) logic this one sum up the whole of his ideas on the matter and demonstrates a stunning adequacy to experimental results from cognitive psychologist James Hampton.
To quickly highlight the main point, it amounts to have the concepts somehow warp their "expected" combination (conjunctive or disjunctive) depending on the instance member to be tested for.
To this purpose Diederik Aerts deploys the full gear of Quantum Field Theory with excellent results, however...
Beside being quite an overkill this monstrous apparatus is still in need of a little help from the experimenter in the form of "appropriately choosen" quantum angles which parametrize the sought for quantum interferences.
With such finely hand tuned values Aerts quantum model of concepts combination gets an astounding zero discrepancy with respect to almost all experimental measurements of human responses.
Not only is it highly suspicious to have a "perfect match" with psychological experiments necessarily subject to measurement errors but, alas, a few unruly sample cases refuse to bow to the quantum model dictums no matter the choice of the "quantum angles", what a pity!
This otherwise outstanding piece of work suffers from a common disease, trying to forcefully shoehorn the reality into a whimsical model instead of trying to unravel more cogent causes for the observations.
From both the succesful cases (even if a bit cheated) and the failing ones it should be obvious that: - Whenever concepts are combined in a composite query they interact in intricate "non classical" ways.
- Quantum logic as such isn't the perfect match to model this interaction.
Why the heck would it have to be?
Like the Platonic Solids were to match the planets orbits?
That was by itself an excellent enlightening exercise in new concepts formation, but now would somebody please give us the actual model of concepts interactions which show up in Hampton's experiments?
 Submitted by Kevembuangga [/html] KevembuanggaonThursday 06 November 2008 - 19:42:35   comment: 3 Thursday 15 May 2008 The smell of Strong AI In my decades long survey of the field I had totally overlooked the works of Jerry Hobbs, yet he seems on the right path, still a bit entangled in "ontology problems" may be and having too much interest in the silly Semantic Web (I hope it's only a matter of budgets...) but I think he is definitely tackling the right questions at the meeting point of Language and Logic.
 Submitted by Kevembuangga
KevembuanggaonThursday 15 May 2008 - 19:04:44 comment: 0 Sunday 06 January 2008 Roger Schank defectsIn a recent answer to The Edge Annual Question Roger Schank said:
When reporters interviewed me in the 70's and 80's about the possibilities for Artificial Intelligence I would always say that we would have machines that are as smart as we are within my lifetime. It seemed a safe answer since no one could ever tell me I was wrong. But I no longer believe that will happen. Of course, GOFAI won't do it!
The wrong headed directions promoted by the biggest stars of AI have done much, much harm to AI, even leading (in 1984 already!) to a panel discussion at AAAI-84: The Dark Ages of AI where Schank himself acknowledged the problem, even later elaborating on his own view of the difficulties he said:
It is not that AI needs definition, it is more that AI needs substance Wrong!
AI does need a definition!
No wonder that this didn't spur any improvment in the field, the culprit was the obsessive focus on logic that he now recognize in the Edge answer :
Early AI workers sought out intelligent behaviors to focus on, like chess or problem solving, and tried to build machines that could equal human beings in those same endeavors. While this was an understandable approach it was, in retrospect, wrong-headed.
Yet, he doesn't really grasp why this was wrong-headed, it's not because "Chess playing is not really a typical intelligent human activity".
He hints at the real cause ("How can we imitate what humans are doing when humans don't know what they are doing when they do it?") but doesn't quite come thru to the full conclusion:
We need to know what we want to do, no just tinker around with "promising" research projects.
The various current crazes about natural language processing, robotics and even statistical learning will be as deceptive as GOFAI, may be bringing a few high-tech gimmicks on the side but not of more decisive import than expert-systems which were the main outcome of GOFAI.
The common trap in which all AI researchers seem to fall is that they are always working on already human-elaborated concepts and whatever "good tricks" they find for processing and "massaging" those concepts more efficiently they are oblivious to the fact that they made up the concepts at start not the computer.
The most basic question therefore seems to be what do we put in a concept?
When and how do we come up with the words we use to talk and think, apple, water, red, line, cosine, mountain, etc...
In the rest of the Edge answer Roger Schank basically give up.
Well, good riddance sir...
One less nearly useless budget eater!
 Submitted by Kevembuangga KevembuanggaonSunday 06 January 2008 - 17:48:16 comment: 1 Monday 24 December 2007 Seeing the forest from the trees (of "open problems") Anyone having had an interest in AI for a while has surely collected a list of "open problems" or "key problems" like the one Peter Turney recently blogged about.
I haven't made an exact tally of mine but among the 2000-3000 papers for which I kept a record there are probably over 50 which I can certainly call "key problems", so far so good then what the heck do I do with this now?
Pick one which I deem among the most importants and try to "work on it"?
Mmmmm...
How likely is it that I will do better than the original researchers?
I would need to have an edge over them on something.
Which "thing"?
Some of these guys have spent a lifetime on their "favorite problem" and I am not even sure I really grasp the matter they are delving, though I can spot it as an important point.
Some are utterly brilliant Matthew Brand, Dominic Widdows, Kenny Easwaran (only a random sample from memory, those are a dime a dozen...)
Some are both utterly brilliant and spent a lifetime on hard problems Grothendieck, Feferman.
Am I (or almost anyone else) really hoping to "magically" make a breakthrough where those people seem to be lingering over?
Yet, yet...
What they are all doing is tackling the trees, some of those among the hardest trees.
For reasons of competitive research (previous post) they have to produce evidence of their smartness and hard work.
This is some kind of depth first search, a hit or miss, the most brilliant get some nuggets the less brilliant or just infortunate get nada.
In any case this doesn't shed much light over the whole landscape, the forest...
What else could be done, instead of forcefully digging even deeper?
I see two options:
- prioritizing the problems, looking for the one(s) which are likely to have precedence over others in the (hypothetically assumed) path toward solutions.
- reframing one or more know problems into a different question by "stepping back" and trying to find a more general framework within which the pending problems will be seen as only a special case of a more general question (going "meta", see about Jacques PITRAT, the only usefull reference I found since he is not an english speaker).
As usual I "just have" to do it...
Submitted by Kevembuangga KevembuanggaonMonday 24 December 2007 - 17:57:20 comment: 0 Thursday 20 December 2007 Is competition a problem?Daniel Lemire and Peter Turney have had an argument about the benefits of competition versus cooperation in science.
Since my opinions are (as usual...) no too well received on other people blogs I will summarize here what else I see as problematic with competition.
In a succeding post of Peter Turney he mentioned that Einstein was inspired by a book by Henri Poincaré, to me this is a perfect example of what Daniel Lemire said :
"After all, being the first to solve a given scientific problem, is important.
Science is a winner-takes-all game, at least some of time."
Actually it can be said that Einstein was "inspired" by the works of many (Poincaré, Lorentz, Minkowski, Hilbert, Grossmann, ...) though his groundbreaking publications did not include references to the work of others.
And after a few productive years (roughly 1905-1915) Einstein's findings appear remarkably bland and very much lacking of any insight ("God doesn't play dice" re quantum mechanics...).
Puting aside the vexing nature of this for the forgotten contributors and not to detract to the performance of having been able to make a synthesis of research trends of his time, this also breeds among the public a lot of misconceptions about science, lone geniuses and the like, which are used by a few scoundrels to sell snake oil or politically loaded agendas, "miracle cures" for this or that, creationism, toxic cults, etc...
So whenever Peter Turney says :
"There is a conflict between competition and cooperation in science.
I feel that conflict myself, but I strive for cooperation.
I have never regretted sharing my ideas."
May be he is overlooking a few other nefarious effects of competition beyond the selfless/selfish antagonism.
But even the misrepresentaion of science within the general public doesn't appear to me as the most serious problem with science.
What seems much more dangerous is the end result of giving the "competitive naked apes" a lot of powerfull gizmos which they tend to use as carelessly as if they were stone axes.
Breaking the neighbour's skulls with a stone axe is of course unfortunate but doesn't entail much damage, having Kim Jong-il, GW Bush and Ahmadinejad playing "nuke poker" is a game in a wholly different league (it seems to me...).
I do not share the optimism Peter Turney shows in his "Second Most Important Research Problem", I see no good reason to suppose that cooperation can be agreed upon in most cases because "private" interests whether of groups or individuals are likely to always conflict on one point or another.
There is no such thing as the "general good" and it is not a matter of lack of rationality in trying to define it.
This is why though I find AI a very valuable research goal I am not really at ease with it.
Not because, like the silly paranoid Singularitarians (Hanson, Yudkowsky, Anissimov, etc...), I fear that the "Big Bad Autonomous AI" will take over humans and cull the masses of useless apes, but because the said useless apes are very well able to turn any technical capability into "improved" means to pursue their lethal competitives practices.
Because in the end it's not the "intelligent monkeys" who are in control but the ones with the biggest egos and the "biggest balls".
Submitted by Kevembuangga KevembuanggaonThursday 20 December 2007 - 10:02:39 comment: 0
News Categories
| |
[news] Almost right...
