Will AI Change — or Eliminate — the Mind-Body Problem?Can an artificial intelligence program that calculates really understand mathematics? Or is that a “hard ceiling” for AI?
In last week’s podcast, Walter Bradley Center director Robert J. Marks interviewed Concordia University philosopher Angus Menuge on the difficult mind–body problem: Dr. Menuge sees mind–body interaction as a transmission of information between two realms; our minds and bodies are one integrated system with a translation function… like developing and then writing down an idea. But what about artificial intelligence? We are told that artificial general intelligence (AGI) is now pushing towards a machine that can totally duplicate the functions of the human mind. But what if the mind is not simply a mechanical function of the brain? What if it is non-algorithmic and non-computable?
This portion begins at 29:04 min. A partial transcript, Show Notes, and Additional Resources follow.
Robert J. Marks: I still have one more question that I want to ask you. If indeed dualism is true, doesn’t that mean that we will never be able to have artificial general intelligence where we have a strict duplication of human performance?
Angus Menuge (pictured): Yeah, I think it does. There will be artificial general intelligence in the sense that there are very sophisticated learning algorithms that can generalize, and so they can move from their initial training domain to work in new areas. So at the level of just being able to formally solve problems, you could say there’ll be artificial general intelligence. However, what you’re asking about is will it really duplicate everything about the human mind? And there I think, no, because I don’t see any reason from these amazing enhancements of the complexity of these systems to think that the systems would move from not having subjective awareness to having it or from moving to true intentionality about anything beyond themselves.
So I think that the fundamental issues are metaphysical. We’re aware that there’s something it’s like to be us and that we can think about the world. And we can also think about things which, it is arguable, no physical system ought to be able to think about — abstract principles, like the laws of logic or theorems about prime numbers. Well, no physical system has ever physically interacted with any of these things. So the very contents of our thoughts seem to suggest that we have access to a realm. In a way it’s a somewhat platonic realm, but without getting into that issue, that’s certainly a realm of things which are not purely physical.
We know, for example, lots of things about the set of integers. There’s an infinite number of integers. We can prove theorems, for example, by mathematical induction that apply to every one of them but all physical causal interactions seem to be finite. How then can an AI physical system ever get to the point where it can truly be said to understand or know things about these sets?
Yes, it will be able to follow through rules that will come out with the right output that agrees with the mathematician’s output. This is true. But I don’t think it can be said really to understand what an infinite set is or what prime numbers are.
Note: Integers are whole numbers — positive, negative or 0, but not fractions. An infinite set is a set of numbers “containing an uncountable or infinite number of elements.” – Story of Mathematics The natural numbers from 1 on are an infinite set. No natural number is so large that we cannot just add +1 to it. And +1 to that one as well. So it is infinite because there is no natural limit. A prime number is “a whole number that cannot be made by multiplying other whole numbers,” like 2, 3, 5, and 7. – Story of Mathematics
Robert J. Marks (pictured): Even on the most fundamental level, a computer can add the numbers two and three, but it has no understanding of what the numbers two and three are nor does it really understand addition. It can do the operation but has no understanding of what’s going on. And so I agree with you. I don’t think artificial general intelligence, where we have a strict duplication — not mimicking, I think mimicking is possible — but a strict duplication of human performance, is possible. I think that is a hard ceiling for artificial intelligence.
Here are the earlier parts of the series:
Part 1: How do we know we are not just physical bodies? The mind–body problem is one of the most difficult issues in modern philosophy. Philosopher Angus Menuge cites the immateriality and indivisibility of the mind and discusses the evidence from near-death experiences.
Part 2: If the mind and body are so different, how can they interact? A look at different models of the mind–body problem. Angus Menuge asks, Why should wanting a drink of milk produce physical changes like opening the fridge? It’s a harder question than many think.
Part 3: How have various thinkers tried to solve the mind–body problem? Philosopher Angus Menuge explains why traditional physicalism (the mind is just what the brain does) doesn’t really work. Some philosophers today claim that the mind is simply what the brain does; a newer group thinks the mind emerges from the brain but is not simply the brain.
Part 4: How would Angus Menuge resolve the mind–body problem? From his background in computer science, he sees mind–body interaction as a transmission of information between two realms
Menuge argues that our minds and bodies are one integrated system with a translation function … like developing and then writing down an idea.
- 01:12 | Introducing Dr. Angus Menuge, professor and chair of Concordia University’s philosophy department
- 04:06 | What is the mind-body problem?
- 06:37 | Near-death experiences
- 10:05 | The history of the mind-body problem
- 15:24 | Popular mind-body problem models discussed today
- 18:14 | Epiphenomenal thoughts
- 22:31 | Dr. Menuge’s take on the mind-body problem
- 29:04 | Will artificial intelligence ever be able to duplicate the functions of a human?
- Dr. Angus Menuge at Concordia University
- The Inherence of Human Dignity, vol. 1: Foundations of Human Dignity edited by Dr. Angus Menuge
- The Inherence of Human Dignity, vol. 2: Law and Religious Liberty, edited by Dr. Angus Menuge
- Religious Liberty and the Law, edited by Dr. Angus Menuge
- The Blackwell Companion to Substance Dualism, co-edited by Dr. Angus Menuge
- Rene Descartes, French philosopher and mathematician
- Aristotle, Greek philosopher
- Thomas Aquinas, 13th century philosopher and Catholic priest
- Thomas Hobbes, 17th century English philosopher
- Jaegwon Kim, Korean-American philosopher
- Richard Swinburne, professor of philosophy at Oxford University
- The Illusion of Conscious Will by Daniel Wagner