Hacker News Reader

Comment by krakensden | Hacker News Reader

If you assume that our brains are completely physical objects (no soul), then you should be able to model them with a computer, like all other physical objects, and create a thinking computer.

Just because we have an incomplete understanding now doesn't mean we always will, or that it's impossible to have a complete (or reasonably complete) understanding.

Or there's the idea of a non-physical, impossible to model soul. I don't see much data to support that conclusion though, it seems the equivalent of cosmic ether to me.

View on HN

Those are not the only options.

See: http://en.wikipedia.org/wiki/Quantum_mind and http://en.wikipedia.org/wiki/The_Emperor%27s_New_Mind

Both of those arguments in no way counter the notion that the brain, as a physical device, can be simulated mechanically- they merely suggest that a simulation would have to be more fine-grained than neuroscience currently accepts. This is much like arguing that since Go has a massively larger game tree than Chess, writing a program to play Go as well as a human is impossible. Intractable? Maybe. Impossible? Clearly not.

It's also worth quoting the second-to-last sentence of the second article you posted:

"The book's thesis is considered erroneous by experts in the fields of philosophy, computer science, and robotics."

It's not that there are "data" to support the idea, but there is the fact that consciousness is unobservable. That is weird and different from all other physical phenomena, wouldn't you say?

Unobservable? You mean it doesn't emit visible light, right?

It's complex, it's hard to define, but it's definitely not unobservable. Also, as a mere physical manifestation, you can fiddle with it in a pretty physical fashion. You can put in on drugs and watch it warping. You can toy with its physical structure surgically and see how it gets messed up (wonderfully discussed in Oliver Sacks' books), you can bump it with a boxing glove and literally switch it off.

Too much interactive for an unobservable phenomenon.

Sorry I forgot the qualification: excepting one's own experience of one's own consciousness. That was silly.

What makes it weird and different from physical phenomena is the fact that it is not generally observable, even though it is posited to exist for all people.

Or do you disagree even with that? If you do: I don't just mean that it doesn't emit visible light. I mean it doesn't emit anything. Consciousness is just the fact that the universe exists from a perspective. It is not an object, "out there", "in the world", but something only experienced by each consciousness for itself. The mere fact that we could create a perfectly convincing AI and then still have doubts about whether there really is "anything in there" is a result of the fact that consciousness is unobservable. Does that explain well enough what I mean?

I'm sorry, but your argument still seems diffuse to me. "Not an object", "anything in there"... these are terms so vague that anything we say about them could stand as irrefutable.

Your own feelings about your consciousness are unimportant to define consciousness. It's definitely irrelevant that you can't "experience" how other people feel the colour yellow, we still regard it as observable. The only difference is that we can define "yellow" as electromagnetic radiation with certain wavelengths, whereas we'll have to wait a bit to get an equally satisfactory definition for consciousness.

We shouldn't attribute a special epistemological status to the brain/mind/consciousness just because it's complex and we don't know it very well and just because it's ours.

If we're to assume that consciousness only exists where it is observable, only I exist. All of you are automatons.

In fact, because of the fact that it is unobservable, you cannot even make a reasonable argument that the netbook I'm typing this on lacks consciousness.

I for example can argue that anything that is executing a series of algorithms as part of a system has consciousness, and therefore this computer is quite conscious, just as I am (even though my series of algorithms is much less well-defined and reproducible.)

I personally do not assume that consciousness only exists where it is observable. I am merely interested in the fact that it is unobservable-- and in how that makes it weird and different. Do you agree with me on this point?

Yes, but in that it's empirically indistinguishable, from an ethical standpoint we should assume it exists when we see the hallmarks of consciousness (though perhaps the primary hallmark is an intrinsic understanding of the concept itself.)

Of course something that does not really exist is unobservable. Consciousness is just an illusion.

Reply on news.ycombinator.com