h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Intelligence not found here, or I wouldn't have been daft enough to even consider starting work on this subject. I started after the fuss and bother when http://www.bbc.co.uk/h2g2/guide/U110017 Playboy Reporter posted his entry on it in the http://www.bbc.co.uk/h2g2/guide/F59234?thread=116555&skip=0&show=20 which resulted in his leaving h2g2, hopefuly not for good. I therefore undertook to take over this entry using the helpfull stuff on that thread, and produce a set of entries which will hopefully be suitable for inclusion in the guide. I am creating this thread so that discussion of the more general stuff does not clutter up the threads for the individual entries. I am also going to include credits to various people here without which this would never have been attempted. these include: http://www.bbc.co.uk/h2g2/guide/U173889 Arpeggio http://www.bbc.co.uk/h2g2/guide/U160070 Barton http://www.bbc.co.uk/h2g2/guide/U116149 Bright Blue Shorts http://www.bbc.co.uk/h2g2/guide/U99875 Mikey the humming mouse http://www.bbc.co.uk/h2g2/guide/U110017 Playboy Reporter who started the ball rolling and of course myself for being daft enough to pick up the ball after he threw it away. If I have missed anyone, then please point it out and I will try to make sure you get the credit you deserve.

The first set of pages are now readable, but not complete, and as it became obvious that this was a multi-page project, it has been submitted to the university.

The pages are:

http://www.bbc.co.uk/h2g2/guide/A582220 - specific intelligences
http://www.bbc.co.uk/h2g2/guide/A582239 - general intelligence
http://www.bbc.co.uk/h2g2/guide/A582257 - intelligence tests
http://www.bbc.co.uk/h2g2/guide/A582266 - what is intelligence
http://www.bbc.co.uk/h2g2/guide/A582284 - machines and intelligence

Thre are also some pages that have been created to bookmark the names, so that they can be linked to before they have been writen. These are:

http://www.bbc.co.uk/h2g2/guide/A582248 - evoked potential
http://www.bbc.co.uk/h2g2/guide/A582275 - animal intelligence
http://www.bbc.co.uk/h2g2/guide/A582293 - intelligence and genius
http://www.bbc.co.uk/h2g2/guide/A582301 - theories of intelligence

and judging from the 86K of notes that are currently there, plus other reference material as well, there may be a few more.

You seem to be breaking things out nicely, but there obvious categories you have not gotten to yet which are mentioned in that notorious 96k. So, I won't mention them again here except that I suspect you are going to find that pretty much everything breaks down to categories of pre-Piaget, post-Piaget, pro-Piaget, and anti-Piaget. These are just my uninformed opinions from very general reading. I am not an expert in this area, though I have very strong reactions against certain factions.

One thing I think you should consider right away though is your use of the Russian AI standard: someone or something is intelligent if it gives intelligent answers to intelligent questions. As you have stated it, this is a useless definition since it is seemingly self-referential. I think that it is necessary to write an article on this standard which carefully defines the the three usages of 'intelligent' in the definition as to actual meaning and scope of application. As it stands, I see little distinction between the Russian test and the Touring test. You evidently do. So give, already!

Barton

The turing test is to do with the gulability of people, which is only partially resolved by turning it into the turing measure. The russian definition carefully makes things like searle's chinese room arguament look as silly as it is, because under the russian definition, the room + operator combination is intelligent if it is acting intelligently, whereas searle seems to think that the operator can be intelligent, the room can never be intelligent, and the combination is therefore not intelligent. Another problem with the turing test is that it only applies for testing when the people you have competing with the machine are within a certain narrow intelligence band.
I will try and find somewhere to put the piaget related stuff, as well as the clarification given above, but I thought it better to get something out there for comment so that I could start getting feedback rather than just working from my interpretation of the 86k of stuff.

I guess I had always 'assumed' that it would be someone intelligent asking intelligent questions and expecting intelligent answers for the Turing test and I'm somewhat suprised that anyone would have taken it any other way.

For me the difference is that the Touring test is a subjective and relative measurment. That is, I might think test indicated intelligence after asking about whether the man in the moon should be forced to wear formal clothing and receiving an answer that the man in the moon is a woman and she'll wear whatever she wants for that reason. Whereas, you might decide it failed because you asked if God could really have expected us to accept so many different churches and were answered that God is only dog spelled backwards and were told, further, that religious questions aren't suitable topics for the test.

Each of us would have made subjective decisions based on relative judgments of success.

However, good will in the direction of the Touring test can clear that problem up by insisting on a series of tests normalized for 'human' response in the manner of the MMPT.

The Russians seem to be striving for a more objective standard but their definition is totally dependant on having an objective standard for intelligence in the sense of an intelligent response to an intelligent question. Ultimately, they are applying a least common denominator standard to the Touring Test and implying that they have objective standards which the Touring Test does not. That is why I need to see what the definitions of 'intelligent response' and 'intelligent question' are since that will suggest what they consider intelligence to be. If they haven't any such standard then their definition of AI is no more valid than Tourings and, in some ways, is worse since they by implication do not allow for human judgement.

Consider the Eliza program. Many people can sit down at the terminal and talk to Eliza for hours and come away feeling that they have truly been helped. Many might swear that the program must be intelligent. Yet, I or anyone else can sit down at that terminal and show that the program is simply throwing out pseudo-random responses according to easily deducible rules and that there is no 'intelligent' response if you violate those rules. I did this myself before I ever had a chance to view the code.

But what are we supposed to say about Shrdlu, this program can answer rather complex questions about the world of blocks it understands and can manipulate with a mechanical arm (or the same in a virtual world.) It is capable of following extremely complex instructions relating to concepts of placement, color, and size which also require it to take initive in generating other moves which have to be inferred as being necessary from the required commands. In fact, it behaves essentially the same as a person would under similar conditions.

Here's a sample session with Shrdlu:

Person: PICK UP A BIG RED BLOCK.
Computer: OK. (does it)
Person: GRASP THE PYRAMID.
Computer: I DON'T UNDERSTAND WHICH PYRAMID YOU MEAN.
Person: FIND A BLOCK WHICH IS TALLER THAN THE ONE YOU ARE HOLDING AND PUT IT INTO THE BOX.
Computer: BY "IT", I ASSUME YOU MEAN THE BLOCK WHICH IS TALLER THAN THE ONE I AM HOLDING.
Computer: OK. (does it)
Person: WHAT DOES THE BOX CONTAIN?
Computer: THE BLUE PYRAMID AND THE BLUE BLOCK.
Person: WHAT IS THE PYRAMID SUPPORTED BY?
Computer: THE BOX.
Person: HOW MANY BLOCKS ARE NOT IN THE BOX?
Computer: FOUR OF THEM.
Person: IS AT LEAST ONE OF THEM NARROWER THAN THE ONE WHICH I TOLD YOU TO PICK UP?
Computer: YES, THE RED CUBE.
Person: IS IT SUPPORTED?
Computer: YES, BY THE TABLE.
Person: CAN THE TABLE PICK UP BLOCKS?
Computer: NO.
Person: CAN A PYRAMID BE SUPPORTED BY A BLOCK?
Computer: YES.
Person: CAN A PYRAMID SUPPORT A PYRAMID?
Computer: I DON'T KNOW.
Person: STACK UP TWO PYRAMIDS.
Computer: I CAN'T.
Person: WILL YOU PLEASE STACK UP BOTH OF THE RED BLOCKS AND EITHER A GREEN CUBE OR A PYRAMID?
Computer: OK. (does it)
Person: WHICH CUBE IS SITTING ON THE TABLE?
Computer: THE LARGE GREEN ONE WHICH SUPPORTS THE RED PYRAMID.
Person: IS THERE A LARGE BLOCK BEHIND A PYRAMID?
Computer: YES, THREE OF THEM: A LARGE RED ONE, A LARGE GREEN CUBE, AND THE BLUE ONE.
Person: PUT A SMALL ONE ONTO THE GREEN CUBE WHICH SUPPORTS A PYRAMID.
Computer: OK. (does it)
Person: PUT THE LITTLEST PYRAMID ON TOP OF IT.
Computer: OK. (does it)
Person: DOES THE SHORTEST THING THE TALLEST PYRAMID'S SUPPORT SUPPORTS SUPPORT ANYTHING GREEN?
Computer: YES, THE GREEN PYRAMID.
Person: WHAT COLOR IS IT?
Computer: BY "IT", I ASSUME YOU MEAN THE SHORTEST THING THE TALLEST PYRAMID'S SUPPORT SUPPORTS.
Computer: RED
Person: HOW MANY THINGS ARE ON TOP OF GREEN CUBES?
Computer: I'M NOT SURE WHAT YOU MEAN BY "ON TOP OF "IN THE PHRASE "ON TOP OF GREEN CUBES " .
DO YOU MEAN:
1 - DIRECTLY ON THE SURFACE
2 - ANWHERE ON TOP OF ?
Person: 2
Computer: THREE OF THEM.
Person: HAD YOU TOUCHED ANY PYRAMID BEFORE YOU PUT THE GREEN ONE ON THE LITTLE CUBE?
Computer: YES, THE GREEN ONE.
Person: WHEN DID YOU PICK IT UP?
Computer: WHILE I WAS STACKING UP THE RED CUBE, A LARGE RED BLOCK AND A LARGE GREEN CUBE.
Person: WHY?
Computer: TO GET RID OF IT.
Person: WHY DID YOU DO THAT?
Computer: TO CLEAN OFF THE RED CUBE.
Person: WHY DID YOU CLEAR OFF THAT CUBE?
Computer: TO PUT IT ON A LARGE GREEN CUBE.
Person: WHY DID YOU DO THAT?
Computer: TO STACK UP THE RED CUBE , A LARGE RED BLOCK AND A LARGE GREEN CUBE.
Person: WHY DID YOU DO THAT?
Computer: BECAUSE YOU ASKED ME TO.
Person: HOW DID YOU DO IT?
Computer: BY PUTTING A LARGE RED BLOCK ON THE TABLE ; THEN LETTING GO OF IT ; THEN PUTTING A LARGE GREEN CUBE ON IT ; THEN LETTING GO OF THAT CUBE ; THEN PUTTING THE RED CUBE ON THAT CUBE; THEN LETTING GO OF THAT CUBE.
Person: HOW MANY OBJECTS DID YOU TOUCH WHILE YOU WERE DOING IT?
Computer: FOUR OF THEM.
Person: WHAT DID THE RED CUBE SUPPORT BEFORE YOU STARTED TO CLEAN IT OFF?
Computer: THE GREEN PYRAMID.
Person: THERE WERE FIVE BLOCKS TO THE LEFT OF THE BOX THEN.
Computer: NO, ONLY FOUR OF THEM: THE RED CUBE, TWO LARGE GREEN CUBES AND A LARGE RED BLOCK.
Person: PUT THE BLUE PYRAMID ON THE BLOCK IN THE BOX.
Computer: OK. (does it)
Person: IS THERE ANYTHING WHICH IS BIGGER THAN EVERY PYRAMID BUT IS NOT AS WIDE AS THE THING THAT SUPPORTS IT?
Computer: YES, THE BLUE BLOCK.
Person: DOES A STEEPLE
Computer interrupting: SORRY, I DON'T KNOW THE WORD "STEEPLE".
Person: A "STEEPLE" IS A STACK WHICH CONTAINS TWO GREEN CUBES AND A PYRAMID.
Computer: I UNDERSTAND.
Person: ARE THERE ANY STEEPLES NOW?
Computer: NO.
Person: BUILD ONE.
Computer: OK. (does it)
Person: CALL THE BIGGEST BLOCK "SUPERBLOCK".
Computer: OK.
Person: HAVE YOU PICKED SUPERBLOCK UP SINCE WE BEGAN?
Computer: YES.
Person: WHY DID YOU DROP IT?
Computer: BY "IT", I ASSUME YOU MEAN THE SUPERBLOCK.
I CAN'T EXPLAIN A NON-EXISTENT EVENT.
Person: IS THERE ANYTHING TO THE RIGHT OF THE RED PYRAMID?
Computer: YES, FOUR THINGS: THE BOX , THE RED CUBE, THE BLUE PYRAMID AND THE BLOCK NAMED SUPERBLOCK
Person: THANK YOU.
Computer: YOU'RE WELCOME!

Why isn't Shrdlu intelligent? It responds intelligently to intelligent questions and commands.

It is not intelligent because, whatever else might be the case, the definition of intelligence will not be restricted to just dealing with a bunch of blocks. And, therefore, will fail both the Russian and Touring tests, though neither, specifically, specifies breadth of intelligence.

Barton

Oops, make that the MPPT (Minnesota Polyphasic Personality Test)

Barton

I agree, shrdlu is not intelligent, but if you had 1000 shrdlu's each dealing with their own area of expertise and another one that deals with integrating it, you might be slightly less sure that it isn't.
Note, because of the way that the statement was phrased by the russians, we are talking about how you define intelligence, not about if the computer can ever be intelligent. This in and of itself is an advance over the turing test, as no matter how well any machine passed the test in it's are of expertise, there are some people who would not accept that it was doing anything that was remotely cever. It also rules out some simple cases like eliza by definition, thus adding further clarity to the debate. For this reason, I intend to stick with the russian definition in the article untill someone here is daft enough to stick their head over the parapet with a better one.
Keep up the good feedback. is the art of the nearly impossible.

NEWS: I am getting positive noises from tptb about it being a university project, so I have created a project index page over on http://www.bbc.co.uk/h2g2/guide/A584525 and it might be an idea to move conversations over to there. Please keep tuned for tomorrows exiting episode of "is he daft enough to keep pluginng away at it until it is finished" (I know, it can never be finished, it is too controversial).

You really don't need to finish it. You just need to touch all the bases. Unfortunately, they keep adding bases to the game.

I think a quick survey is possible but I also think that a university project is the better choice.

It wouldn't hurt to ask some of the people who were commenting on the original article if they would be interested in contributing something on the aspects of intelligence they were commenting on, particularly if you specify that what you need is just a quick paragraph or ten to cover that particular point.

Please note, I didn't comment on the original article.

Barton

Incidently, 'possible' should have been preceeded by the word 'barely,' but, you will note, I did not feel competant to try and was not interested in becoming competant enough to have a go at it. I am glad that you decided to take up the gage. What you have done so far, entitles you to take the 'stupid' out of your screen name. I think you can brazen it out with the word 'ignorant' which after all is a curable condition and one which you are rapidly working to remedy. Something that I have the utmost respect for.

Baton (who doesn't have the time, right now, to stop being ignorant on the topic of intelligence but who hopes that you will help him cheat his way there)

A first draft of entries are now waiting feedback over at the university project on http://www.bbc.co.uk/h2g2/guide/A584525 where any help would be greatly received.

I have attached a few quick comments to the articles.

Barton

This is not intended negatively, but having just clicked through the Intelligence entries I'm finding they are nothing like my idea of what they might look like.

I applaud your efforts to try and bring an end to the bickering that was going on, but still feel there is a long way to go.

With the greatest of respect may I suggest you obtain a copy of an "Introductory Psychology" textbook from your local lending library and reading the chapters on Intelligence within it. If you have already done this then I apologise. This is not meant rudely, only to help. A good book might be that written by "Malim & Birch" or "Gross".

I realise that psychology is not the be all and end all to intelligence, but I think that as it has spent over 100 years trying to investigate this phenomena (which can never be absolutely proven) it deserves to be used as the basis to such entries.

Unfortunately I do not have time to contribute properly, or I would do so. Therefore maybe my whining is inappropriate.

Anyway good luck
BBS

as regards using textbooks, the following sources are being used, in addition to the contributions from h2g2'ers:

check your own I.Q. by H.J.Eysenck has some good stuff in the introduction,

modern psychometrics - the science of psychological assessment by John Rust and Susan Golombok is filled with stuff about testing, sources of inacuracy, etc

Human Intelligence - its nature and assessment by H.J.Buther has lots of stuff about intelligence

Science and sanity by count alfred korzybski

I have also read a lot about the mind, covering psychology, computational modeling, philosophy, general semantics, nexialism, sociology, cultural morphology, and I could name quite a few other subject areas as well.

About the only thing that the different areas agree on (certainly not terminology) is that they don't know what intelligence is, how it works, why it works, and no two definintions of what it is are even slightly compatable (even in the broad details) when you starts looking in detail at them.

They do agree that learning has to take place to extend the reach of an intelligences world view, but if you are satisfied with a very intelligent being who has already learnt enough to get by in the world, you don't even need the learning.

Given that, I did not want to swallow whole the terminology from any single discipline, and thus thought it better to try and come up with my own definitions and usages that while not identical to any of them is broadly compatably with all of them. (if you disagree withany of them I will discus that particular definition of the project page).

This will no doubt annoy any psychologist who looks at it and can be as pedantic as barton (note to barton: sorry, but you have described yourself that way on more than one ocasion, so I thought that I might get away wth it in this instance), but if I didn't, I would get an equal amount of flack not just from psychiatrists, but psychologists, neurologists, neurobiologists, etc (the list could go on but you get the idea).

Yep, I'm very pedantic in certain situations.

I am particularly so when it comes to trying to find answers to questions that are easily obscured by misunderstandings due to sematics. There is no point in having a discussion turn into an arguement simply because the parties involved are using the same words to talk about different things or even using different words to talk about the same thing.

The modern practice is to identify the key concepts early in the discussion and carefully define terms to have the specific qualities of those concepts. These ad hoc definitions are then prominently stated and identified at the start of the discussion then those terms are used with exclusively those meanings by all.

In a paper this is a straight forward fashion. In a discussion, however, the moderator has the responsibility of enforcing that useage and attempting to interpret new ideas in terms of the vocabulary of the project.

(Was that pedantic enough for you? )

So, since you say you have tried to invent a vocabulary for this discussion, you must first prominently publish those definitions using terms that clearly atributable to other speific disciplines or deliberately avoid using the terms of other disciplines. Then you must scrupulously use your new vocabulary with cautions in the text when you might be misunderstood by someone from some particular field if you can possibly anticipate such issues.

So, I guesss I'm looking for a new article about definitions of the terms of this discussion. This will be your safest article since it will all be correct by definition. Though some may feel that your vocabulary is incomplete and others may feel that some term is defined in such a way that it might obscure a point that they feei is important. While such complaints are valid and should be answered, they cannot really be discussed until all the areas that that term is pertinent in are finished.

Besides it feel *soooo* good to be able to say to someone, "Oh! You haven't understood our definition. Have a look at the definitions page so that you will be able to understand what we are saying here."
It's not being mean, its recognizing the basis for misunderstanding is simply a question of terminology that has already been addressed.

Good work so far! You are clearly becoming our local expert in a nexial and hyper-synthetic fashion on as many areas of study and influence as you can delve into. Keep it up!

Barton

I try my best, but I am not good enough yet at the nexialism to class myself as anything but an enthusiastic amatuer.
As regards the definitions, I might have to create a definitions page, but at the moment, I am trying to define on first use, and make all the pages interlink properly with each other, and with the "intelligence" overview which I haven't been able to write yet.
hopefully, I can sneak the definitions into the overview, but I need suggestions for how to do it.
As so many disciplines use the same words to talk about the same things, but with definitions framed in incompatible ways, I am trying to keep the usages similar, while trying to use different definitions that are broadly compatable with multiple disciplines. By trying to declare on first use, in the most apropriate place, and linking it all together into a naxialistic whole, I hope to sneak the non-standard definitions through with nobody noticing and thereby create a standard(ish) vocabulary for the edited guide that no-one will strongly disagree with. If later specialities can then be added, the action of trying to link them in will sneak the (new standard) definitions into the people doing the writing, and hopefully try to extend this vocabulary further.
What do you think, can I get away with it?

Nope!

But, I really don't see a need for trickery. There is no harm in defining a set of terms that seem to work. How to get them in?

Something like:

Vocabulary

Since there are many different ways of approaching the study of intelligence, a similar number of specialized vocabularies (or jargons) have sprung to better describe what is being discussed. Unfortunately, many of the words used in these vocabularies are the same even though they are used to describe different things or different approaches. In order to avoid you having to learn all these different terminologies, these articles will use the following vocabulary which is general in nature to help explain what intelligence appears to be. When necessary, distinctions between these definitions and the others used will be noted in the text or a footnote.

List

New Heading


There's nothing quite like the direct approach. Using this method allows you to directly integrate objections and exceptions to your definitions into footnotes where appropriate.

And, I think you are wise to put this stuff in the body of the introduction/overview. You may end up with an article on vocabulary anyway for the sake of clarity. I suspect that need will become obvious before you're done.

Barton

Hi all~


We're feeling exceptionally vacuous this early morning, with yet another emergency run to hospital to get hooked up to an oxygen machine -- this time at home -- until the lungs get a grip on themselves and stop forgetting the trick. The lungs went on strike because of an appallingly high mould-count in the environment. Stupid lungs. Another necessary evil, just like indispensible but infuriating people.

Note to self: never have a medical emergency at the same time the cops are hauling the sorry arses of two rival gangs, who were busted during a war, into hospital in droves and handcuffs. They'll put you on oxygen and forget you for several hours once you're not going to die.

We are not sure, on the whole, that creating *yet another* synthetic meta-jargon is the best approach. It might be the only approach. This is something we are going to re-re-read the articles and consider, because *something* is bothering us, but we haven't nailed down what it is, yet. This morning would not be a good time to try to figure that out.

BTW, Barton, in slinging our way through here, we noticed a trivial correct remark you made, which you later emended: it is the Minnesota Multiphasic Personality Inventory, or MMPI, as you first said. If the damn thing is calibrated to measure humans, this is the first we've heard of it . That stupid test is so antiquated it makes a mockery of all psychodiagnosis. We have never met a reputable psychodiagnostician who would use the MMPI in isolation and call the results significant unless they were downright *strange* -- and only people who understand the instrument, or are thoroughly insane (there is quite a lot of overlap; we've met reputable psychodiagnosticians, but not too many pleasant ones, let alone sane) can skew the results that heavily. MMPI is a good argument against any reliable, universally applicable, quantifying tool as applied to human personality.

Personality informs (to use the jargon, because I persist in doing that, to everyone else's annoyance) intelligence, to at least the same extent as intelligence informs personality -- more likely a lot more (or this would be a boring, pedantic, OFF-topic post, and not just a boring, pedantic one). There are huge, vast, positively terrifying numbers of people whose personalities succeeding in conquering and subsuming their intelligence. Some of them are right here at h2g2. (No, Xyroth, it is not just you... ) How many people can anyone claim to know, whose intelligence *alone* has subsumed their true personalities? Concealed, sure. Any sociopath at the high end of the bell-curve can do that. Beaten down until the personality appears to be gone... verrrry unlikely.

On the other hand, personality disorders can be concomittant with, be caused by, but do not create intelligence. (Omitting long and complicated discussion of 'awareness' of 'alternate realities' at the moment.)

On the gripping-hand, no one has successfuly proven that the chicken precedes the egg, or vice-versa. Any theorist who argues the 'tabula rasa' non-intelligent life angle (Freud to Georg Simmel, oy!) about babies obviously never was one. Infant human intelligence is almost not studied at all, because the babies and poop their nappies, and generally are lousy at answering even the simplest Likert-type questionnaire. They also get passive-aggressive, and go to sleep. As such they are annoying to study, so very few people do. So they are generally accused of 'instinct', 'mimickry', and occasionally credited with 'emotion'. This is dismissive, but a self-supporting, self-reinforcing science like cognitive development or psychodiagnostics finds infants *impossible* to describe.

Behaviourists have no trouble with this little roadblock at all.

Our thinking is trying to deteriorate on us... Does O2 normally do that? It never did, at sea-level. Well, we really only came here to create a link and to say, no Xyroth, you are not imagining anything. If you're finished with your ba-ba now, perhaps one of the ACEs can take it back to Momma. No . Momma don't like it when you . The usual term is 'infantilisation'.

Kassandra, prating on as usual, for LeKZ

It seems to us that the whole issue of how different disciplines employ different terms would make an awesome entry, by our local Jargon Pundit, Barton (as usual, Barton, if you or any of your team is injured or captured, the IM Force will deny any knowledge of your activities. This tape will self-destruct in five seconds).

All right. That'll be enough of that, Kassandra. I'll make suggestions but I will not write an article for this project. If I thought I had something creative rather than reactive to add I would have volunteered.



Barton

B~



~K/L