MindForth Programming Journal (MFPJ) 2010 August 17

Tues.17.AUG.2010 -- Using Gender to Trigger Who-Queries
Today we would like to see if the AI
can ask a who-query rather than a default
what-query, if the gender of a noun in
question is known to be masculine or
feminine. In English, as opposed to
German or Russian, a non-neuter gender
indicates that an entity is a "who"
and not simply a "what".
When we rename 11aug10A.F as 17aug10A.F
and run the Forthmind, entering just the
word "god" causes the following exchange.
Robot: GOD WHAT IS GOD GOD
Human:
Next in the AskUser module we insert a
diagnostic message to reveal any value
held in the "mfn" gender variable.
Robot: GOD
AskU: mfn = 0 WHAT IS GOD GOD
Robot: GOD WHAT IS GOD GOD
Human:
Apparently any value that may have been
held in "mfn" for "GOD" has been reset to
zero by the time the AskUser module is called.
We should be able to run a ".psi" report and
check for sure. Oops! We chose the wrong report.
We run the ".en" report.
324 100 0 1 1 100 5 100 322 to GOD
329 101 0 0 0 101 2 101 326 to HERE
333 102 0 0 1 102 5 102 331 to MAN
339 103 0 0 0 103 5 103 335 to MEDIA
346 104 0 0 0 104 5 104 341 to PERSON
352 105 0 0 0 105 2 105 348 to THERE
357 106 0 0 0 106 7 106 354 to WHOM
363 107 0 0 2 107 5 107 359 to WOMAN
367 56 0 0 0 56 7 50 365 to YOU
371 67 0 0 0 67 8 58 369 to ARE
380 108 0 0 0 108 5 108 376 to MAGIC
383 58 0 0 0 58 8 58 382 to BE
389 100 0 0 1 100 5 100 386 to GOD
393 100 0 0 1 100 5 100 390 to GOD
398 54 0 0 3 54 7 54 394 to WHAT
401 66 0 2 0 66 8 58 399 to IS
405 100 0 0 1 100 5 100 402 to GOD
409 100 0 0 1 100 5 100 406 to GOD
t nen act num mfn fex pos fin aud
The above ".en" report on the English
lexical array is encouraging, because it
shows that the word "GOD" retains its "mfn"
value of one (1) for masculine each time
that the word "GOD" is used. However, the
software may be blanking out the "mfn"
value in advance of the AskUser module.
We need to run a search on "mfn" in the
formatting link
code to see in what situations the
"mfn" value is reset to zero. Hmm, "mfn"
is reset to zero after storage in the
InStantiate module. In order not to disturb
the extremely fundamental InStantiate
functionality, we should perhaps create
"mfnflag" as a variable to pass the gender
information from InStantiate to the AskUser
module.
Tues.17.AUG.2010 -- Post-Upload Upshot
We did create and use "mfnflag" to get the
AI to ask "Who" when a noun had a male or
female gender, but not without some difficulty.
We were coding under time- pressure, and the
new "mfnflag" kept losing its value somewhere
between its initial setting in the InStantiate
module and its utilization in the WhoBe module,
but we could not at first detect that the value
of the "mfnflag" was being changed -- probably
by the occurrence of a zero-gender word like
"WHO" itself. Our fix was to protect the "mfnflag"
value within an IF-THEN clause in the Instantiate
module, so that the positive value of "1" for
male or "2" for female would persist until
dealt with in the WhoBe module. Unfortunately,
such a quick fix may be less than ideal for
many normal situations.
It is typical of our AI coding that we latch
onto even a sub-optimal algorithm that proves
our point, so that we can get the functionality
up and running. We were in such a hurry that
we tested the AI only by entering the word "god"
and seeing our desired response of
"GOD WHO IS GOD" and not "GOD WHAT IS GOD".
Maybe right now we will test the AI to see if
it reaches the fourth call to ReJuvenate and
then properly asks, "GOD WHO IS GOD".
We tested the 17aug10A.F AI and we let it run
through the four activand concepts of KbTraversal.
When it activated the concept of God, it said
first "GOD WHO IS" and then "GOD WHO IS GOD",
so there are still some bugs to be worked out.
The AI also said, "I WHO IS AM I", which is a
step backwards in functionality. On the whole,
however, the AI is approaching self-referential
thought.
We will need to firm up strongly the concept of
self or "I",
Reply to
Mentifex
Loading thread data ...
Huh?
I could more or less agree that processing natural language well means intelligence. (Hofstaedter). An acquaintance of mine, a linguist, had some ideas and has had help of a programmer in Java. I've seen a demonstration of his analysis of a small story, and it really is able to identify objects in the story. (Really identify in the sense that the ``her'' on the last line, is arguably related to ``girl'' in the first line.) It has some inroads to understanding where a new topic starts, and even where an old topic is picked up. This may be world class, state of the art research etc. but your attempts are so pathetic compared to this.
The difference is probably that he doesn't refuse to stand on the shoulders of giants.
Groetjes Albert
--
Reply to
Albert van der Horst
I there a published version of the program which recognises that "her" refers to the girl. I have though of ways of tagging pronouns and finding out what their noun is. One method is to work backwards and find the first noun which agrees in number and gender. I think something more sophisticated must be hapenning.
- Ian Parker
Reply to
Ian Parker
Unlikely: "Mentifex" is a well known crackpot.
His "achievements", such as they are, are approximately equivalent to the level of Terry Winograd's "shrldu", (written something like thirty years ago); but he posts each minor change as though it were the birth of Skynet, and never responds to other posts.
In other works, he's a crackpot and a troll: ignore him.
Reply to
David Mitchell
Definitely. Those linguists have collected properties for all words in the English language. For a sentence like "Jesus is god" they can build on a database that knows that god is a substantivum, a person, male, unconcrete etc. The program I've seen correctly identifies subjects, objects verbs etc. in non-trivial sentences, then builds from there. That is not to say that it could make sense of the incoherent crap that Mendifex feeds his baby with.
His ongoing work is confidential. I'll see whether I can dig up some references. Meanwhile those linguistic databases are public domain (as they should) and you could use them.
Groetjes Albert
--
Reply to
Albert van der Horst
Wrong on two accounts. The above quote of Ian is not about Mendifex, it was about a program I have seen working.
Secondly, Terry Winograd shrldu stands as one of the great achievements in AI, and ...
Fully agree, but you realise what an insult that is to Winograd?
Groetjes Albert
--
Reply to
Albert van der Horst
Why? Winograd wrote his SHRDLU more than 40 (not 30 as stated above) years ago when it was a breakthrough. Today it isn't, thus writing any comments on someone writing yet another SHRDLU is irrelevant to Winograd.
Consider gliders. Lilienthal built his gliders like around 120 years ago, it was a major breakthrough then. Today, if someone comes up with a "fresh" idea of building a "kite" that flies by himself without bond to the ground, and presents this as a major breakthrough in aviation, one has all rights to call him a crackpot or anything else. That has nothing to do with Lilienthal since he did what noone had done before him. Today one can read almost anything about building a glider in a library, or join aircraft company and build planes. Even as a hobbyist one can build and fly gliders much better than Lilienthal's or aircraft more powerful than Wrigth's Flyer.
Reply to
Aleksej Saushev
Sorry to go off-topic, but I think there is a world of difference between a "glider" and something like Lilienthal's "hang-glider". While Lilienthal made some impressive flights in his weight shift control hang-gliders, it was the Wrights that invented the first glider with 3-axis control in 1902. While weight shift hang-gliders exist yet today they remain athletic toys with no practical use. However today's latest design gliders and even jet airliners still use the 3-axis control principle invented by the Wrights which was the real major breakthrough in flight IMO.
-Doug
Reply to
Doug Hoffman
There are a number of engineering concepts. Is Lift/Drag in the concept dictionary? I could make a list of terms specifically concerned with one topic, say aviation.
The word "better" to me is L/D + wing loading a measure of stall.
Heck why write
formatting link
70 is absolutely amazing. At 1km (3,300') you can glide 70km in still air - amazing. The table given should give green aviators pause for thought. At over 20:1 it is greener to fly than to drive a car. Dick Rutan has achieved great things.
But getting back to concepts. If I am designing an aircraft I use certain concepts. I have found from experience that linguists are all arts trained. If you go into a technical field like aviation, the concepts involved do not seem to be there. CP Snow talked about 2 cultures.
What are the concepts associated with "stall". In aviation it means minimum flight speed. The speed at which increasing the angle of attack reduces lift. It means different things in different contexts. A full context, arts linguists note includes aviation.
A concept dictionary and a Von Neumann context free language is vital for MOLTO
formatting link
gay in English is the equivalent of "lustige" in German. Google Translate translates everything through English and so we got "homosexuel" as the translation (correct was "lustige"). MOLTO will need a concept language. MOLTO aims to provide quality translation between any language pair. It will have its source code available. Whether this includes a concept dictionary I don't know.
Existing dictionaries like the Oxford English dictionary will provide a concept dictionary. You need to wade through it though. Have MOLTO requested any help from the OED team?
What you are working on is exactly what MOLTO will need. I just wonder how far people have got both in MOLTO and elsewhere.
- Ian Parker
Reply to
Ian Parker
His programs are utterly trivial compared to shrdlu. Putting them in the same league again is an insult to Winograd.
I respect a hobbyist who successfully builds a glider much more than I do Mendaxfex. His program would not be a contender in a Loebner contest.
Reproducing the level of AI of shrdlu in Forth would be a highly respectable achievement.
Groetjes Albert
--
Reply to
Albert van der Horst
If something has performance better than indicated in Wiki, it is a true breakthough. Mind I believe the challenge for aircraft designers, particularly light aircraft designers is to achieve performance closer to what Rutan has shown is possible.
The Cessna for example is a very old design. Anything would be a breakthrough compared with that. Wiki as I have indicated expresses the state of the current art.
In the same way Google Translate is "State of Art". When we point out genders in Hebrew, for example. Google has decided that driving a car and playing football ALWAYS has the male form of the verb. Catering for the many women drivers Israel has, and their women's football team would be a breakthrough therefore.
Google BTW has been arrogant in their responses.
formatting link
You must define state of art. SOA is that in French we get "Le magazin" (shop) - OK. Elle fut =E9tabli. It can't find the gender correctly. They are different sentences. Still working backwards "Le magazin" is the first singular noun. This defines a kind of SOA.
- Ian Parker
Reply to
Ian Parker
That really emphasized Aleksej's point: Lilienthal's weight-shift glider was a tremendous achievement when he did it, but today there's room for argument about whether it should just be called a "glider" or have a qualifier.
Reply to
Joe Pfeiffer
Well, OK, he *hasn't* managed to recreate 40 year old AI concepts...
Reply to
Joe Pfeiffer
The Wright brothers developed warping surfaces as a means of controlling an aircraft. Lilienthal' used weight distribution. This clearly would not work in a large powered aircraft. They dis gain patents for the technique.
Stability in aircraft is in fact an interesting point. Birds are unstable, all aircraft up to the F16 were stable. The F16 was the first aircraft to be provided with a fly by wire system. Unstable aircraft have an advantage in lift/drag. On Concorde the tail plane was pressing down, dragging and hindering lift. On the F16 the tailplane lifts. You need a computer to control an F16 which is why it was not done before.
In recent times alternatives to ailerons and elevators (Wright's warping) have been devised. Drill holes in a wing and such air in. This technique reduces the moving surfaces need. It also provides the potential of a method of controlling turbulence. The ins and outs of this, particularly in modern aviation, are quite interesting. Holes in the wings can delay stall and can thus replace flaps.
I really ought to cross post this in rec.aviation.military
- Ian Parker
Reply to
Ian Parker
The wing warping in essence was just another way to install ailerons. The major part of the warping occurred at the trailing edge of the tips, just like ailerons. Wright wing warping = ailerons = roll control only. The Wrights did not use warping for pitch or yaw control. Instead they invented and used conventional elevator and rudder.
They even identified the advantages of a longer wing (higher aspect ratio) for improved glide (L/D ). They also identified the problem of adverse yaw which led to the movable rudder to counteract, just like most aircraft today especially long-winged gliders.
The lack of control is what killed Lilienthal. The Wrights died of non-aviation related causes.
Sure, lots of variations can be used. But I would estimate that the vast majority of new aircraft built yet today use something that closely resembles the Wright's 3-axis control (with ailerons doing essentially the same job as wing warping).
-Doug
Reply to
Doug Hoffman
The vast majority do indeed. Chord ratio is an interesting point. Indeed a high chard ratio will increase L/D at high speeds, at low speeds thick wings provide more total lift.
One incentive for looking for novel control methods is UAVs, particularly at the smaller end. If you have something the size of a large insect there is an incentive to investigate novel control systems. Another incentives is that holes in the wing are being used to reduce turbulence. Turbulence is in essence chaos. If you are reducing turbulence you might as well also get rid of ailerons and elevators. Computers have in fact increased the range of options.
- Ian Parker
BTW - In rec.aviation.military most of the contributors are very conservative and many do not understand the fundamental principles.
Reply to
Ian Parker
Hi, Regarding the Mindforth, project, I don't know enough to offer an overall critique.
I like the idea of multi-year projects that explore a single, continuous line of research, as I believe that this benefits our field tremendously.
Presumably those who snipe at this project have taken a long, deep look at the code and the underlying ideas.
My main concern or question about this line of research is that, offhand, it seems to me that much of the knowledge of the IA (intelligent agent) is hardwired into code. I'd like to see an explanation by the developer of underlying (declarative) knowledge representations, if any. Or at least some notion that the IA can modify its own knowledge and/or "learn" in some fashion.
Rules about the use of "I", "me", the verb to-be, and so on seem to be stuck into code, rather than developed as part of a more flexible representation.
Even so, I respect the notion that someone is pursuing a long term project along whatever path, as much is learned along the way. This contrasts sharply with most research (in the US at least), that is year-to-year, hand-to-mouth, survivie-to-the-next-funding-cycle, and so on.
Regards, Amnon
Reply to
amnon.meyers
Take a look at
formatting link
It would, were he as good a programmer as he claims to be, but he's not. The main talent he displays in public is self-promotion.
I did, many years ago when I first encountered him. All I remember about it was a profound sense of disbelief that anyone would claim that it represented any kind of breakthrough whatsoever, as it seemed trivial and obvious.
Not by Arthur - he's still banging the same nail he started with over twenty years ago (and, amusingly, claimed was 99% finished in June 1999).
He has nothing to teach anyone else in the field either, so it's pretty much been a total waste of time.
Reply to
David Mitchell
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
To me the projects we should be looking at are the following.
MOLTO - This is the EU language program. I have already referenced this. I would like to know in more detail how they are attacking multiple language translation without preferring one language as GT does (English). A Von Neumann language that would be implied by MOLTO's objectives is indeed the first step to AGI.
Web 3.0 - This implies the ability to put in links automatically. MOLTO will in fact produce links.
Mizar The basis of Mizar is Tarski-Grothendieck set theory.
formatting link
OpenMath
formatting link
OpenMath is now convertible to Mizar
formatting link
Open CYC which deals with atoms
All these things are interconnected. I would like to see something which would interface up and connect to the above.
- Ian Parker
Reply to
Ian Parker
I feel I should add a few things.
1) Software engineers are adamant that the language is unimportant. For this reason I would like to see things written in the more mainstream languages like C+or Java. In fact in C++/Java the "class" can be used as a concept designator.
2) I have brought in Mathematics. In fact MOLTO aims to be able to translte mathematical statements. As AI consists in putting concepts into a formal basis all AI (ultimately) must be presentable in OpenMath. I must be able to do this with any statement even something like Politics which one does not associate with mathematics. I don't know whether MOLTO are aiming to base their "Grundsprache" on OpenMath.
3) The EU CANNOT give any language (except Mathematics and a generated Grundsprache) priority. GT gives English priority, everything is translated into English first.
- Ian Parker
Reply to
Ian Parker

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.