25 Years of the AAAI
Posted 11 Jul 2005 at 20:35 UTC by steve
article notes the achievements and "tremendous progress" of AI
since the first meeting of the American
Association for Artificial Intelligence 25 year ago. Of course, the
article notes, despite that progress, "machines still aren't smart
enough to do just about anything a person can". AI proponents point out
examples of AI including chess playing programs, search
engines, and the cruise control in your car. The usual prediction of
intelligent machines within 10 years won't be found in this article
though. The experts
asked predict it will be a "long time" before we achieve real AI.
As posted at http://wedesoft.blogspot.com
Hello, Jan Wedekind. I have just read your Amazon AI4U
review and I feel that you have been very unfair. My work is far
different from SHRDLU. You
write that I should withdraw my theories if
proven wrong (and I agree with you), but who has proven me wrong?
People have ridiculed my theory of mind, but they have not pointed out
any particular part of the theory that is wrong, and they have not
explained why anything in the theory is wrong. On the contrary, people
like the eminent AI researcher Dr. Ben Goertzel
express admiration for the Mentifex AI theory. I am sorry if you bought
my AI4U book and did not like it or perhaps did not understand it, but
I hope that you will hold on to the AI4U book and come to appreciate it
more as time goes by. I am not *pushing* people to buy the book; it is
available to read for free on-line. If you eventually
realize that your review is wrong, I hope that *you* will retract your
review. If my AI theory is proven wrong, I will retract it, but thus
far it has not been proven wrong, even after quite a few years. In
closing, this message is a friendly comment. I appreciate your taking
the trouble to write a real-name review, and even if I disagree with
what you say, I will defend to the death your right to say it.
WoW!, posted 15 Jul 2005 at 16:35 UTC by dogsbody_d »
I'm impressed M. That was an actual post, not just a link. Okay, so it
was cross-posted, but even so...
There's a maxim in writing er... well Dr Who fiction, but I'm sure it's
general... Don't tell, show. So if your theories are so fantastic,
knock it up into a robot (how long can it take ;) ) I applaud your open
attitude, but I don't think it's the science that puts people off, it's
the attitude. I really did think that you were a bot for a while!
So, do us a favour, and try, just try, not to post a link to your site
for a while, but KEEP POSTING. I don't care if people think you are a
loon, they'll say I'm a crackpot too. Albeit a crackpot with an MSc in
AI. Crackpottery is no reason for ignoring you. Speak up I say! Just
try not to speak up with such blatant spam.
Please post real posts, what do you actually think about the news item
you are responding to? Does it remind you of anything similar? Come
on, you've been bothering the field of AI since before most of us could
walk, you must have some experience!
P.S., posted 15 Jul 2005 at 16:35 UTC by dogsbody_d »
Talking about yourself in the third-person is weird.
I just love the AI4U FAQ
that wedesoft posted!
You know the creation of any product involves many steps such as
theory, prototype, test, rework, produce product and maintenance.
Without being able follow all such steps and working only on the
theory part it's like a total waste of time. And a mind is a terrible
thing to waste! AI4U has never gotten past vague theories! AI4U
doesn't have any proof of concept, only a glob of random, hard to
follow, ideas that he thinks a mind should be. Hey, even I have ideas
how a mind works, but so what, that doesn't make me know for sure how
artificial intelligence works as he claims. He doesn't know how to
assemble his thoughts into a working artificial intelligence or even
understands for sure lowly bug intelligence. So how can he be an
expert in the mind? Why believe anything he says until he shows proof
that it can do something useful? He should put up or shut up which
from seeing the years this guy has been at it, I don't expect he will
shut up even though he should.
ATTACKING THE SPURIOUS-ASSOCIATION PROBLEM
It should not matter what inputs we use as a test to debug the problem
of spurious associations. It would be nice, however, to find some
classic inputs that would highlight any existing problem.
For test input, if we use words that are already in the AI bootstrap
vocabulary, then NEWCONCEPT should
not get called and the AI should generate a response based solely upon
the functions of OLDCONCEPT, ACTIVATE and SPREADACT.
However, when we type in
"people see robots"
"PEOPLE SEE YOU"
as a response. Our
diagnostic mode reveals to us that the spurious direct-object "YOU" had
an activation of 54, while the correct direct-object "ROBOTS" had a
high but insufficient activation of only 51.
Uh-oh. We have a vexing enigma of a bug right now. With the test input
above, we only get the wrong response in diagnostic mode, not in normal
mode. It suggests a Heisenbergian problem where to observe the
functionality is to change the functionality.
After much experimentation based on guesswork, it seems that the
#56 psi concept of "YOU" is being activated whenever we start the Forthmind
either the space-bar or the Tab key.
Hmm. The spurious activation of the "YOU" concept has something
to do with the bootstrap "ME" concept that went just before it. Oh,
well, it's not so much a spurious-association problem as it is a
spurious-instantiation problem. In fact, dwelling on the problem leads
us to speculate, if not conclude, that the AI is just doing what it is
supposed to do after the unterminated input of a SPACE character, which
is to instantiate the current psi concept. Therefore the inner-POV
concept of "ME" gets instantiated as an external-POV concept of "YOU".
After all, the Forthmind
is in a
mode of accepting external input.
Ideas for a fix come to mind, such as somehow not letting the Tab
key lead to a calling of INSTANTIATE
This session of coding reminds us of how "brittle" the AI
software is. We introduce the Tab key functionality, and we break the
I often lose my mind too when I can't get my TAB or CAPS key to work.
Sometimes, I look down and INITIATE a key press of the CAPS key and it
doesn't work all the time because of the cola fluid that I
accidentally SPILT down the keyboard and haven't had the TIME to run
it through the dish WASHER. So, when that happens I become INFURIATED
and then want to YANK on my hair or what I have left of it until I
become BALD. So, instead, I just go FORTH and buy another keyboard
and save my MIND but so now my operating system still works a little
off KILTER but my keyboard is OK for the time BEING.
... from someone as busy as Mr. Murray.
> ... and I feel that you have been very unfair.
I rather think, that you're pretending to feel like this for
> My work is far different from SHRDLU.
That's my point.
> ... (and I agree with you), but who has proven me
There is no need to prove it wrong, because it's obvious. The burden of
the proof is lasting on you.
> People have ridiculed my theory of mind, ...
They will continue to do so, as long as you don't stick to the rules of
the scientific community.
>...but they have not pointed out any particular part of the
theory that is wrong, and they have not explained why anything in the
theory is wrong.
People have done this in length already. For the moment you've forfeited
the right to get reasonable explanations.
>... Dr. Ben Goertzel express admiration ...
No, he doesn't (see AI4U-FAQ).
>I am sorry if you bought my AI4U book and did not like it
It's not a matter of liking (and please don't remind me of my mistake).
> ... or perhaps did not understand it, ...
What is more likely?
>If you eventually realize that your review is wrong, I hope
that *you* will retract your review. ...
In the unlikely case, that you'll forget about this book and come up
with some proper stuff (*), I'll consider the book as one of your
>In closing, this message is a friendly comment. I appreciate
your taking the trouble ... defend to the death your right to say
Don't even think about my right to say it! And now you can safely
dissect my posting, because I'm waiting for (*) to happen.
Mentifex said at..., posted 18 Jul 2005 at 23:19 UTC by AI4U »
The future is
not so bright
according to Rolling Stone
will arrive first? Vernor Vinge's Technological Singularity
based on artificial
, or The Long
where oil depletion causes the collapse of civilization?
When M.I.T. (all those technology squares) recently held a
("Hi! When do you come from?"), nobody showed up from
the future to report that AAAI-05
finished AI graduate school
and had gone
on to save the human race via AI. What went wrong in the future?
Perhaps the problem is the great AI schism between Academia
and as-it-were the Impressionists of AI, the motley crew of AI hackers,
phrackers, also-rans and Singularity fans. You of the AI Establishment
hold your convention in the cathedral, while out in the bazaar, the bizarre
organized an avant-garde Salon des
. When I get to the future, I am not coming back, so please
listen to what Mentifex
has to say right now.
If you wonder when will the
-- it is happening *now*. A flashpoint is
approaching where resistance will turn to acceptance
antagonism will turn to enthusiasm
Currently the muggles ascribe the darkest of motives to your
humble independent AI scholar trying to spread the AI word. Take for
- which is free to read
on the Web. On Amazon, the
ludicrous Luddites post AI4U
reviews with such titles as
rantings of a crackpot
. If you prick us, do we not bleed? And
if you poison us, do we not die? Therefore, when we respond to you
somebody please write an
that sets the record straight?
I'm really happy to talk to people outside of academia. I think most of
us are. The robot community happily embraces those people who produce
results. Where would we be without linux? The BEAM community? Robot
Wars has encouraged a whole generation of little roboticist kids.
Various competitions, from micro-mouse to robocup have brought people
together from all over what is a very multidisciplinary field. We've
got computer scientists, mechanical engineers, electronics people,
philosophers, psychologists, operational researchers, ethologists,
entymologists, and ethicists. Oh, and you.
For goodness sake, meet people halfway. I'm not going to tell you to
obey the rules of scientific community maaaaan, but a little netiquette
goes a long way. So I note that you've crossposted AGAIN. As well as
netiquette, you also violate a fairly normal principle of language. If
you talk about yourself in the first person all the time, you sound like
a nutter. Go on, reply o this post saying "I am not a nutter." If you
say "Mentiflex is not a nutter" it won't count.
Okay, rambly-nut boy. Good to see you providing citations for your
claims, but seriously, Rolling Stone? Rolling bloody Stone? Not Time,
or Newsweek, or New Scientist, or anything? Sorry we couldn't be
bothered to come back from the future for you, but that conference was
for time-travellers, thus relying on TIME TRAVEL being invented/legal
etc. so it's more their problem than the AI community's.
Nice reference to the Salons btw. Dunno where you got that idea, but
as I was saying, we like all of those folks. Quote from the Wiki "Most
were poor quality, leading to ridicule in the press." Where do you fit
in I wonder. Isn't it true that you piss off most of the bazaar too?
Opinion varies as to when/if the singularity will happen, but it really
isn't now. Not in any way that makes any sense. Infinite technological
expansion? Do you know something about the PS3 that we don't?
Oh, we are so not Luddites. I mean, duh! Oh, and spamming AI
forums does not constitute spreading the word of AI. We've heard the
word, we've even heard your word. What we've not really seen are your
actions. Mark Tilden was a maverick. He was an outsider. What did he
do? Spam and troll and talk about himself in the third person, or even
first person plural? Nope, he ended up with a shed-load of cash and
being the darling of the robotics community 'cos of Robosapiens. Can't
you get your software to do something useful, instead of bullying
us to do it all the time.
I ascribe no dark motives to you. My only concern is that you might
actually be mental. It's my background that means that I care :) I'm
worried about you dear. You never actually seem to really engage with
people in conversations on the internet, I only hope you have real life