Comments on: Is the brain too strange to emulate?

By: Will Brown

Will Brown — Tue, 11 Sep 2007 02:06:55 +0000

Ahh, I think I see the problem. A Turing machine is programmable for independent operation, but not necessarily able to pass the Turing test for independent intelligence.

Discovering whether or not I can pass the test can be left for another occasion.

By: Stephen Gordon

Stephen Gordon — Mon, 10 Sep 2007 05:40:24 +0000

Will: Also, check out this Speculist post.

By: Stephen Gordon

Stephen Gordon — Mon, 10 Sep 2007 05:36:45 +0000

Will: Isn't a Turing machine simply an archaic term for AGI? No. Read here about the related concept of "Turing completeness."

By: Phil Bowermaster

Phil Bowermaster — Sun, 09 Sep 2007 22:04:52 +0000

Legion –

It’s not correct to assume that brain function is describable by what most computer scientists think of as an “algorithm.”

Allow me to make a response that is just bas logical and evidence-backed as your position:

“Yes it is.”

Not unless you consider the universe itself to be an algorithm.

Any reason why we shouldn’t? Other than your say-so?

By: Stephen Gordon

Stephen Gordon — Sat, 08 Sep 2007 11:35:21 +0000

Legion:

Follow the “Linkathon” link and you will learn that this series is where people email us with links. I can’t provide a link to the email.

Am I certain that biological brains can be simulated in silicon? Nope. Did the words “I suspect” not clue you in on that? This is my opinion.

How many lifetimes of the universe? Well, let’s give it 100 years. If it doesn’t happen within that time “I’d suspect” that it probably can’t happen for one reason or another.

See Arthur C. Clarke on the danger of saying something is “impossible“.

By: Will Brown

Will Brown — Sat, 08 Sep 2007 10:10:52 +0000

I’m Will Brown; mind your manners, [well-phrased personal attack removed.]

As to your emulation idea Stephen, I think the added levels of complexity inherent to that model works against it. Any model that says, “First, build a Turing machine …” sort of misses the point don’t you think? Isn’t a Turing machine simply an archaic term for AGI?

Learning to emulate specific brain functions will almost certainly be useful to the development process of integrating manufactured enhancement into human brain/mind capability. I think that process the more likely (if somewhat serendipitous) route to independent AGI then the purely mechanical emulation one.

By: legion

legion — Sat, 08 Sep 2007 07:11:18 +0000

As I mentioned in a previous comment, there are limits to abstraction in the real world. No one has succeeded in creating an algorithm that describes how the brain does what it does.

By: legion

legion — Sat, 08 Sep 2007 07:09:55 +0000

So even if the brain is not a Turing machine, it could be emulated by a sufficiently powerful Turing machine. In theory.

Are you certain about that? How many lifetimes of the universe would it take? What are your sources?

It’s not correct to assume that brain function is describable by what most computer scientists think of as an “algorithm.” Not unless you consider the universe itself to be an algorithm.

You mentioned Will Brown, but failed to provide a link. Please observe proper link etiquette, as people are watching.