Intelligent behavior
It’s weird that in a set of articles about AI, I’m only finally getting around to talking about what the field of A.I. research has to say. This article relies the leading AI textbook, part Bible and part encyclopedia: Stuart Russell and Peter Norvig’s Artificial Intelligence: A Modern Approach,
Intelligence
Professional and academic research in AI generally avoids trying to define “intelligence” (mostly, I think, because of the kind of philosophical problems I’m talking about in these articles).
They have crafted careful definitions of what they are trying to do with machines. Russell and Norvig define AI research as “the study and design of rational agents” where a rational agent is anything that (1) has a goal (2) acts in the world to achieve that goal.
These goals are defined by a “utility function” or “objective function”. This is a kind of test that produces a number that tells the program how well it accomplished its goal. As far as the program is concerned, the goal is to increase that number. Machine learning programs know how to change themselves to make that number keep going higher as they keep doing their job over and over, better and better.
For example, a large language model is trained on a (huge) amount of text. The program tries to guess what the next word is going to be. If it guesses right, the number goes up. If it guesses wrong, the number goes down. It does this trillions of times, and every time it guesses right or wrong, it tweaks the numbers in a matrix in a way that makes the number go up more than it goes down.
The takeaway is this: AI research defines “intelligence” in terms of well-defined problems with well-defined solutions.
This definition has several advantages: (1) It’s measurable. (2) It uses the same mathematical language as economics, statistics, and optimization (for example the term “utility function” is from economics and “objective function” is from decision theory. This meant that results from these older, more successful fields could be applied to AI programs. This in particular was what ended the AI Winter in early 2000s and has continued to drive AI’s recent success.
If you think about it, the definition used by AI research isn’t that different than the one used by psychology. Psychometrics defines intelligence as “how well do you do on intelligence tests”, which might seem circular at first glance, however the number that they get (called “Q”) is very useful and reliable, perhaps the most reliable measurement in all of psychology. It correlates with many other important numbers (e.g. the likelihood that a person will succeed in school) and it’s highly heritable (which suggests that it is measuring something that is physically “real”).
Intelligence is Multi-dimensional
The takeaway here is that both AI and psychology define intelligence in terms of problems and solutions, measured with tests that return numbers. That’s it. There isn’t any other scientific definition of “intelligence”.
If we had to say what the overall “intelligence” of a program is, it would have to be a combination of how hard the problems are, how well the program solves them and how many different problems it can solve.
If this is the right way of looking at intelligence, then there is no simple way to put a single scalar number on the total “intelligence” of a thing. The whole idea of a linear scale for intelligence doesn’t make sense. We can only measure the machine on a problem-by-problem basis.
You could define an “index” by combining a number of problems (as an intelligence test does or the various benchmarks currently being use to test AI models), but this only obscures the problem. There are many different ways to put together an index and the theoretical justification of any index is profoundly arguable. Just as there’s no particular problem that is “most important”, there’s also no particular collection of problems that is most important.
Thus, intelligence is multi-dimensional. Every machine (or person) is at different levels of “intelligence” with respect to different problems. There are as many kinds of intelligence as there are kinds of problems. Even the “generality” of a machine’s problem solving abilities is a multi-dimensional convex region in problem-space.
Human-Level Intelligence
“Human-level” intelligence is measured by how well an average human can solve a particular problem — it’s different for every problem. For some problems, computers have outperformed humans 70 or 80 years. For others, they still can’t solve them as well as we can. (But the list of such problems has become rather short and is getting shorter.)
Average-Human-level and Best-Human-Level performance on a problem are interesting from a practical point of view, because this effects real world decisions: businesses and everyone else has to decide at what point they should accomplish a task with a machine rather than an employee.
But this difference doesn’t apply to any of the problems that computers are already better at. For example, the “human level of intelligence” on multiplication is completely uninteresting. There’s no reason to think that it will be any more significant for any of the problems that people are currently better at. Perhaps, it’s possible. But there’s no evidence.
Intelligent Behavior vs. Human Behavior
Russell and Norvig dismiss the Turing Test succinctly:
Aeronautical engineering texts do not define the goal of their field as making “machines that fly so exactly like pigeons that they can fool even other pigeons.”
The purpose of AI research is to build programs that can solve difficult problems, and human simulation is just one of those problems. Intelligence and human simulation are two different things.
Human simulation is an active area of research, because it is useful for user interfaces, customer service, video game characters, works of art and pornography. But these are applications of AI — none of them necessarily require the simulation to have a high level of intelligence.
AI research doesn’t need machines “think like human”. It needs them to solve problems that humans solve by “thinking.” This is a different thing. What we need are machines that can solve problems. Especially problems we can’t solve.
What makes a car useful is how different it is from people. A car doesn’t have legs, it doesn’t run, it just solves problems that people used to solve by running. If it was just like a person, it would be useless.
Intelligent Behavior vs. Subjective Consciousness
In The Paradox of Mary and Mark we talked a little bit about the problem of other minds: there is no way to determine whether or not a machine has subjective consciousness from the outside. There’s nothing stopping you from building a machine that is just acting like it has subjective consciousness and no one could tell the difference.
This means that subjective consciousness is perfectly useless for artificial intelligence. If they built a machine that had subjective conscious experience, no one could tell. Why would they go to all the trouble to build a feature into a program that no one can detect?
Russell and Norvig wrote in 2003:
We agree with Turing—we are interested in creating programs that behave intelligently. The additional project of making them conscious is not one that we are equipped to take on, nor one whose success we would be able to determine.
There are algorithms that implement some of the things that the brain appears to be doing when you experience subjective consciousness, like “inner speech” or the “train of thought” or the way our attention system broadcasts information to the entire brain. In 2021, Russell and Norvig amended their opinion to take these into account.
Individual aspects of consciousness—awareness, self-awareness, attention—can be programmed and can be part of an intelligent machine.
However, their wisecrack about the pigeons still applies; again, we’re not simulating human thinking here, we’re solving problems that humans normally solve by thinking.
(Some neuroscientists and a few philosophers have claimed that these kinds of algorithms or measures just are consciousness. Other philosophers disagree. These theories might mimic how our mental machinery seems to work, but they don’t explain why this machinery feels things when it’s doing this. They don’t explain subjectivity and so aren’t relevant to what we’ve been talking about. Ned Block is reported to have said about one of these theories “You have a theory of something, I am just not sure what it is”.)
The Takeaways
If this is the best definition we have for “intelligence”, then:
- There is no universal “level of intelligence”, no linear scale, that can be used to measure all things.
- There is no hard threshold, built into the universe, that divides the “intelligent” things from the “non-intelligent” things.
- There’s no evidence that human intelligence (that is, the class of problems we are especially good at) is uniquely exceptional or is some kind of important milestone.
- Intelligence has no special relationship with other human properties like subjective consciousness or human sympathy.
This creates problems for several assumptions about the future of AI.
The first shows that “intelligence” can’t be the basis of a chain of being: a linear scale that can be used to classify all things in the universe into “higher” and “lower” beings.
The second shows that the Nishmet, the transcendental moment when machines become suddenly more powerful and dangerous, can’t be something that happens at a particular “level of intelligence.”
And the last two show that “intelligence” can’t be the essential aspect of human beings that makes them different from all other things in the universe and brings all the interesting properties of human beings with it.
CHARLES GILLINGHAM 