Michael Pollan

So I live very close to Silicon Valley here in Berkeley, and it is a consensus opinion of the people who work in that world that AI, artificial intelligences, can be conscious.

They base this on a premise, and it's a huge premise that I don't think we should accept.

The premise is, as you described, that basically the brain is a computer.

And that consciousness is software.

And if you can run it on the brain, which is essentially, in their view, a meat-based computer, you should be able to run it on other substrates, other kinds of machines.

So it's just a matter of figuring out the algorithm of consciousness.

Now, there are a couple problems with this, in my view.

I don't think the brain is a computer.

If you look back in history, you find that whatever the cool, new, state-of-the-art, cutting-edge technology was, it was likened to the brain.

That became the metaphor.

So at a certain point, it was the clock.

It's been looms, mills, telephone switchboards, and now it's computers.

There's a computer scientist who once said, the price of metaphor is eternal vigilance.

Just because something is a metaphor doesn't mean that they're equal in any way.

And I don't think people have been very vigilant about this metaphor.

And they use language.

They talk to us in our language.

So it's not hard to anthropomorphize computers or chatbots.

But I just want to pick apart that metaphor for a minute because I think it's really important to get right.

First of all, computers have a sharp distinction between hardware and software.