Artificial intelligence has learned to probe the minds of other computer
July 31, 2018 4:50 AM   Subscribe

Building neural networks to understand Theory of Mind “Theory of mind is clearly a crucial ability,” for navigating a world full of other minds says Alison Gopnik, a developmental psychologist at the University of California, Berkeley. By about the age of 4, human children understand that the beliefs of another person may diverge from reality, and that those beliefs can be used to predict the person’s future behavior. Some of today’s computers can label facial expressions such as “happy” or “angry”—a skill associated with theory of mind—but they have little understanding of human emotions or what motivates us.

The new project began as an attempt to get humans to understand computers. Many algorithms used by AI aren’t fully written by programmers, but instead rely on the machine “learning” as it sequentially tackles problems. The resulting computer-generated solutions are often black boxes, with algorithms too complex for human insight to penetrate. So Neil Rabinowitz, a research scientist at DeepMind in London, and colleagues created a theory of mind AI called “ToMnet” and had it observe other AIs to see what it could learn about how they work.
posted by RickLiebling (4 comments total) 19 users marked this as a favorite
 
Dr. Gopnik has done some interesting work on the connections between David Hume’s philosophy of mind and certain aspects of Buddhism. I’ll be sure to check this out. Thanks.
posted by Barack Spinoza at 5:02 AM on July 31, 2018 [2 favorites]


Dr. Gopnik has done some interesting work on the connections between David Hume’s philosophy of mind and certain aspects of Buddhism.

How an 18th-Century Philosopher Helped Solve My Midlife Crisis: David Hume, the Buddha, and a search for the Eastern roots of the Western Enlightenment
posted by homunculus at 7:48 AM on July 31, 2018 [4 favorites]


(Thanks! Didn’t have time to link earlier.)
posted by Barack Spinoza at 2:29 PM on July 31, 2018


I'm going to read this later when I can give it the time it deserves, but my first and strongest reaction to the post is that this sort of research is necessary for (but not sufficient) anything approaching strong AI which, I want to emphasize, we're not even remotely close to.

But ToM is absolutely crucial, if not the key. In my opinion, the arms race between more complicated behavior and more reliable theory of mind is where consciousness arises, as this becomes recursive. (That is to say, at some point it becomes inevitable to apply ToM to self as part of the ToM.)
posted by Ivan Fyodorovich at 3:33 PM on July 31, 2018 [1 favorite]


« Older Trolls thought I was a man. That saved me.   |   Behemoth, bully, thief: the English language is... Newer »


This thread has been archived and is closed to new comments