Universal tests of intelligence

Mar 24, 2011

(PhysOrg.com) -- A new intelligence test, which can be taken by any living creature is being developed that will enable comparison of intellect between humans and animals.

Up to now, it has been difficult to determine without the assessment being based around a language - which has made it impossible to determine how intelligent a chimpanzee is compared to a human, or a to a worm.

Associate Professor Dowe, along with Dr. Jose Hernandez-Orallo of the Universitat Politecnica de Valencia in Spain believe that a universal test of intelligence can be devised that goes beyond the current boundaries of language.

The test would employ the method of operant conditioning, where the test subject has to work out the challenge by trial and error.

For example, if the test were noughts and crosses, the individual taking the test, having never seen the game before, would first have to work out that the game is won by getting three in a row on three-by-three grid, before playing it. Correct responses would be rewarded in a manner appropriate for the user – i.e., money for a human or a banana for a chimpanzee.

Participants undertaking the tests who are performing well get upgraded to a harder test, while those performing badly get downgraded to an easier , as measured by the principles of algorithmic information theory. This underlies the Minimum Message Length (MML) principle of machine learning, econometrics and statistical and inductive inference.

Explore further: New insights into eyewitness memory from groundbreaking replication initiative

More information: The paper which sets out the work is currently the most downloaded article in the prestigious A journal, Artificial Intelligence.

Related Stories

On the hunt for universal intelligence

Jan 27, 2011

How do you use a scientific method to measure the intelligence of a human being, an animal, a machine or an extra-terrestrial? So far this has not been possible, but a team of Spanish and Australian researchers ...

In search of machines that play at being human

Oct 14, 2009

Researchers at Carlos III University (Spain) have taken part in an international contest whose objective is to improve artificial intelligence utilized in virtual worlds. The challenge for the participants ...

International research tests language tests

Jun 04, 2009

Unique Queen's University research will shed light on how motivation and anxiety affects language test scores, and on the relationship between the test scores and the social and educational contexts of the tests.

'Jeopardy!' to pit humans against IBM machine

Dec 14, 2010

(AP) -- The game show "Jeopardy!" will pit man versus machine this winter in a competition that will show how successful scientists are in creating a computer that can mimic human intelligence.

Recommended for you

Mother-daughter research team studies severe-weather phobia

Sep 19, 2014

No one likes severe weather, but for some just the thought of a thunderstorm, tornado, hurricane or blizzard can severely affect their lives. When blood pressures spike, individuals obsessively monitor weather forecasts and ...

Study: Pupil size shows reliability of decisions

Sep 18, 2014

Te precision with which people make decisions can be predicted by measuring pupil size before they are presented with any information about the decision, according to a new study published in PLOS Computational Bi ...

User comments : 20

Adjust slider to filter visible comments by rank

Display comments: newest first

stealthc
4.7 / 5 (3) Mar 24, 2011
how many humans do you suppose would fail the test?
random
not rated yet Mar 24, 2011
This would be amazing if it actually works the way they expect it to. I can already see the humorous potential.
Simonsez
1 / 5 (2) Mar 24, 2011
Presumably all would fail it at some point, as each win puts you at a harder test the next time. Were you asking how many would never win a single time?
hush1
5 / 5 (1) Mar 24, 2011
One day, when I fulfill any and all definitions put forth for the word "life", I too, will take this test. Until then, consider me unborn, dead or 'inanimate'.
IvyMike
not rated yet Mar 24, 2011
Physorg has already covered this
http://www.physor...nce.html
trekgeek1
5 / 5 (1) Mar 24, 2011
They have to make sure the games they create are truly unique. If the first game is tic-tac-toe, despite whatever name they do or do not give it, people from countries who play that will have an advantage. You may subconsciously recall a similar game that you've played before, whereas some people may not have encountered that game.
Denis_Chinkevitch
not rated yet Mar 24, 2011
as long as reward is real money I'll do my best :-)
thewhitebear
4.2 / 5 (5) Mar 24, 2011
i think this kind of logic underestimates the range of intelligence that may or may not exist in other life forms. No matter how we design such a test it will inherently measure a human concept of intelligence because humans imagined and created the test.
mrwolfe
5 / 5 (2) Mar 24, 2011
how many humans do you suppose would fail the test?


As an intelligence test, it really isn't a test at all, but an assessment. You can't fail it, you can only achieve a low score.
Nyloc
5 / 5 (5) Mar 24, 2011
Each species has evolved the kind of intelligence that's relevant to it. It would not make sense to test all species using a test to recognize symbols and patterns. Testing an ant may involve it's ability to use smells to negotiate maze terrains. Humans may flunk tests that ants excel in. Dolphins would be better tested in the realm of their primary senses at tasks which they need to perform well in to survive, such as finding food in the dark buried in the sand.

Similarly, humans are social animals and may be better compared to other social species in some tests. Language and symbol tests would best be used to compare species who have evolved to recognize symbols.

I fully expect humans to perform poorly at specific tests which worms excel in. Wouldn't THAT be something!
DamienS
4.3 / 5 (3) Mar 24, 2011
No matter how we design such a test it will inherently measure a human concept of intelligence because humans imagined and created the test.

And they also created the notion of intelligence. There isn't a watertight definition for species, which is also a human categorization, and I don't think you can similarly measure cross species intelligence in an absolute way. What does it mean to say a mackerel is more intelligent than an iguana? Ultimately, all of these tests relate back to us and our biases.
ekim
not rated yet Mar 24, 2011
Testing an ant may involve it's ability to use smells to negotiate maze terrains.

Would a single ant be tested or an entire hive?
Collective intelligence could prove to be very powerful.
IvyMike
not rated yet Mar 25, 2011
the first article's right there in the related stories for gods sake
Cave_Man
1 / 5 (1) Mar 25, 2011
So which is smarter, your schools science teacher or a paramecium?

Or how about a talking parrot vs a clump of dirt.

Better yet why don't we just spend alot of money on prostitutes and say we didn't!

Sorry these studies just bug the crap out of me, you might as well just sit in a room with whatever you are trying to measure intelligence wise like a cat or monkey and just come up with an arbitrary number to categorize intelligence. Because things like prions or RNA are not smart but the intelligence they possess is seemingly infinite. How can a piece of RNA detach from your genome and travel to exactly the right spot and modify move or inject a gene right where it needs to go? thats like trying to get up off your chair, walk 600 miles and suddenly find precisely what you need like a note with your mother phone number on it that you were looking for when you got up.

Yes I'm high as shit right now....go figure.
epsi00
5 / 5 (1) Mar 25, 2011
well, animals in the wild tend to be good at what they do to survive. it's what drives them to solve a given problem, who to get food, or where to build a nest...They show us that, in the wild, they solve, very often efficiently, the problems related to their survival. And we know that in the wild animals don't have the luxury of considering problems that are completely irrelevant to their safety, survival, food etc. They could not care less for crosses in a square or if pluto is or is not a planet. It's irrelevant to them. So failing such a test or getting a low score can only tell us that such a task is not part of their " natural problems to solve " at most.
frajo
5 / 5 (3) Mar 25, 2011
Correct responses would be rewarded in a manner appropriate for the user
This implies an assumption that there is a relation between "correctness" and "rewards".
When I was a child we often played "monopoly". But I never tried to get as much money as possible. I tried to get certain properties which formed some pattern only I would know.

Thus, the player's concept of "correctness" and "reward" does not necessarily equate with the tester's concept. But who is more intelligent, then?
Modernmystic
1 / 5 (2) Mar 25, 2011
Until dolphins can devise a a similar test I guess we're "stuck" with us doing it. Therefore any complaints that such a test would be human biased are irrelevant, and trite.
Cave_Man
3 / 5 (2) Mar 25, 2011
Until dolphins can devise a a similar test I guess we're "stuck" with us doing it. Therefore any complaints that such a test would be human biased are irrelevant, and trite.

City birds already developed a test, listen carefully in the morning when one starts to squawk, then in full view of the bird move your head around in an attempt to locate the source of the squawking you should notice a distinct change in the birds "language" if you are indeed in his field of view.
Au-Pu
not rated yet Mar 25, 2011
If the system for finding the correct responses in this test is not pre-programmed into the computer or the computer program, then I will back the worm.
Moebius
5 / 5 (1) Mar 26, 2011
I think that there are other kinds of genius that this test will not uncover or evaluate. Would this test rate a Mozart a high level genius for example? or Peter Jackson? or John M. Browning? or Holless Wilbur Allen?