Hacker News Reader

Comment by danbruc | Hacker News Reader

If anything, the argument works even better now that you've gone and given yourself sonar: you can talk to me, unlike that dumb bat, so you can describe your experiences to me, and yet I will still have no true understanding of the experience itself.

This underlines the suggestion at the very end, we are lacking tools to properly talk about experiences and maybe we could actually develop them. My inability to describe to you, or probably even better to a blind person, what it is like to see a red object may after all not be fundamental. Maybe there is a way to communicate and in turn truly understand what using the tongue device is like, we just have not yet discovered it.

View on HN

One of the more interesting proposals I was faced with as part of my Cog Sci & AI undergrad was that we do indeed have ways of communicating that subjective/experiential knowledge — knowledge which is otherwise ineffable. Art is one obvious example of this. At its best, art is able to communicate things that a purely propositional language cannot. The same could be said of rituals — participation in a ritual is a formalized way of forcing you to take on specific states of being that convey a kind of knowledge that can only be acquired by embodying it.

This might all sound a little floofy, but it's really not. Think about the deep knowledge conveyed through game simulations. I've never been to space, but thanks to hours and hours of playing Kerbal Space Program, I really do have some sense of what it's like to navigate through space — a knowledge that goes far deeper than if I had only learned all of Kepler's and Newton's laws and equations.

There may be some relationship between that and subjective experience, but it's not really the same thing. You're describing essentially "know how" instead of "know that". But here we're talking about something that doesn't actually seem to be representable in terms of physical

Why is it that there's something it's like to be zukzuk? zukzuk could simply be like a computer program, simply responding to inputs. Or you could simply have an erroneous memory of actually experiencing things (instead, you simply just responded to inputs, one of which was to record an apparently subjective experience). That doesn't seem to solve the problem to me; it's just kicking the can down the road a little.

The distinction between "know how" (aka procedural knowledge) and "know that" (declarative knowledge) is a favourite in cognitive science, but I think it is a bit shallow. I guess the point I'm making here is that "knowing how" is a major part of knowing "what it is like" to be something.

For example, I know how to ride a bike, and that gives me some idea of _what it is like_ to be someone who knows how to ride a bike — an insight I certainly wouldn't have if I only understood the mechanics of bike riding in declarative terms. The fact that I will probably never know what it is like to be a bat has a lot to do with my inability to have the know-how that a bat has (it's very hard to know how to fly and echolocate when you don't have wings nor the appropriate sensory organs). However, I'd argue that an active, deep VR simulation of bat flight might bring you just a little bit closer to having that embodied, subjective knowledge, in a way that studying bat behaviour or anatomy from a purely objective perspective never would.

What it will be like? Series of sounds, which reprogram brain structure to develop image processing networks?

I can't say it sounds plausible.

Think about explaining general relativity, quantum chromodynamics, or just the way a second level cache works to a random human from the early stone age. Probably not an easy task. I am not sure if this is actually a fair analogy, describing what seeing red is like is in some sense very different from explaining the theory of general relativity. But on the other hand the situations are also somewhat similar, the stone age guy just does not posses the required language, probably to a large extend math, and knowledge and is unable to look back on several thousand years of human history and innovation that shaped us. Sure, it looks like a stretch to us that we will ever be able to meaningfully convey the experience of seeing a red object, but I am not convinced that this is a given. Intuitively I certainly agree with you, not going to happen, but maybe that is just a lack of imagination.

As a kind of brute-force proof-of-concept: it could be a series of instructions to poke your brain with an electrified needle of how to invoke the same thoughts.

(It could be that they're different locations in your brain that have to be poked than the locations it would take in the sender's brain, but it's not a fundamental obstacle. Either the sender studies your brain to tell you where you should do it, or instructs you how to figure out the right spots.)

But does it sound square?

Reply on news.ycombinator.com