Reshared post from Justin Kiggins

Sharing a comment from +Andrea Kuszewski's thread:

+Andrea Kuszewski Great article!

You say: "The brain is actually not like a computer; it doesn’t always follow the rules.". This is a curious claim. The brain obviously follows some rules, and they are quite clearly going to be rules that follow both electronic (in the sense of synapses firing) and computational in the sense that it can be given a formal (and indeed, a computable) characterization.

In other words: the brain is quite literally a computer. Turing deals with this mistake explicitly in his original 1950's article, under the section entitled "The Informality of Behavior", and rejects it as an objection to artificial intelligence.

What I think you mean in this article, which are otherwise quite good, is that the human brain doesn't work like "computers", those manufactured devices we buy for a thousand or so dollars, that comes with a screen and a keyboard and such. Human brains obviously don't work anything like those things. This is the public perception of a "computer" as a commodity; it is the impression that a computer is a desktop or a laptop, that a smartphone and a tablet are "small computers", and that there are big scary things called "supercomputers" and who knows what they do. These are what give rise to the the sci fi stereotype that computers only "follow rules", or who have to "break their programming" (usually with an "emotion chip") in order to be a real boy. This is the Pinocchio myth of AI, and it plagues the popular discussion.

But that's not what computers are at all. Computers follow rules, but they are the same kind of organizational rules that all computational systems (including human minds) follow. The computers we interact with and use on a daily basis are mostly special purpose machines built to perform a certain set of operations, but we shouldn't use that bias to draw fundamental lines in the sand dividing us from our machines. The mind is a computer, but it is a special type of computer. It's very important that we figure out what kind of computer it is, and you are absolutely right that we don't know how to build machines that can perform all the complex and wonderful tasks that a human being can perform, but that's a historical feature of our current technology and not a fundamental distinction between computational symbionts.

I talk more of Turing and the popular misconceptions of AI here, if anyone is interested:

Justin Kiggins originally shared this post:

My brain is not like a computer … well, not yet.

In this pair of articles, +Andrea Kuszewski makes a strong case that the differences between brains and computers are largely due to how we've designed computers and how traditional AI has been stuck in this framework.

But if we start building computers like brains (leveraging decades of psychology research) and not, well, "like computers", then the statement that "my brain is like a computer" will become increasingly true.
Part 1:
Part 2:

I, Robopsychologist, Part 1: Why Robots Need Psychologists | The Crux | Discover Magazine

Mind & Brain | artificial intelligence | Andrea Kuszewski is a behavior therapist and consultant, science writer, and robopsychologist at Syntience in San Francisco. She is interested in creativit…

Submit a comment