Computer programs that ace IQ tests
February 16, 2012
Researchers at the Department of Philosophy, Linguistics and Theory of Science at the University of Gothenburg, Sweden, have created a computer program that scored up to 150 on specific portions of an IQ test: identifying patterns in pictures and number sequences.
IQ tests include progressive matrices, which test the ability to see patterns in pictures, and number sequences, which test the ability to see patterns in numbers. The most common math computer programs, such as Mathematica, score below 100 on IQ tests with number sequences.
For Claes Strannegård, researcher at the Department of Philosophy, Linguistics and Theory of Science, this was a reason to try to design “smarter” computer programs. “We’re trying to make programs that can discover the same types of patterns that humans can see,” he says.
Modeling human problem solving
They have integrated a mathematical model that models two types of human-like problem solving. They developed two programs.
One solves progressive matrices. It scores IQ 100 and has the unique ability of being able to solve the problems without having access to any response alternatives.
Another program specializes in number sequences and is able to ace the tests, implying an IQ of at least 150 for that class of problems.
They did not address verbal and other types of problems.
“Our programs are beating the conventional math programs because we are combining mathematics and psychology,” says Strannegård. Our method can potentially be used to identify patterns in any data with a psychological component, such as financial data. But it is not as good at finding patterns in more science-type data, such as weather data, since then the human psyche is not involved.” .
The research group has recently started collaborating with the Department of Psychology at Stockholm University, with a goal to develop new IQ tests with different levels of difficulty.
‘We have developed a pretty good understanding of how the tests work. Now we want to divide them into different levels of difficulty and design new types of tests, which we can then use to design computer programs for people who want to practice their problem solving ability,” says Strannegård.

Comments (9)
by walter
I invariably score poorly on IQ-tests because I find them obsolete.
by winston
“The group has improved the program that specializes in number sequences to the point where it is now able to ace the tests, implying an IQ of at least 150.”
This is really bad journalism, and this statement highlights why this article is misleading to the lay reader. Although IQ tests do include mathematical reasoning portions, they also include a whole list of other types of measures that a computer would measure really poorly at. Do not be mislead to believe their metric of 150 translates to a 150 on an *actual* IQ exam.
… In other news, computer program specializing in English vocabulary definitions scores a 200 IQ.
by The SHM
This does not necessarily mean there is any lreal ogic BEHIND this intelligence. A computer, as of yet, is still, as Michio Kaku once said, “an adding machine.”
A real IQ test for a computer would be for it (or, if it passes, him/her?) to be given a test with material is isn’t even programmed to understand and still pass. Still, a great leap forward (or perhaps just a big step) to genuine artificial intelligence
by Robert
As a member of Mensa, I welcome our new silicon-based members.
by Alexander
Because of psychology. If you are very dumb the answer is wrong, if you are smart the answer is right (any answer)
by andmar74
“’1, 2, …, what comes next? Most people would say 3, but it could also be a repeating sequence like 1, 2, 1 or a doubling sequence like 1, 2, 4″
This is ridiculous. If I answer 1,2,1 on an IQ test, it’s wrong ?
by nfordkrz
The point is to give the MOST LIKELY answer. By your own statement, most people would say 3, so that is the correct answer. Language (and even math) is ambiguous. To communicate with others, we must be able to figure out what they most likely mean and respond accordingly.
Even math is ambiguous. What is 1+1? It depends on what the numbers refer to. Nevertheless, if you put anything other than 2 on an IQ test, you are being obtuse.
by Brian
Yes. See Alexander’s answer for explanation.
by melajara
Yes, it is well known that there is an infinity of progressions compatible with the presented subset, the point of the test is to focus on the presentation of the “simplest” or “most obvious” correct answer and this is here where a model of Human Psychology is relevant in the IQ test calibration. Ironically, too clever people could score badly, because normal “Human Psychology” becomes irrelevant for them.