Hacker News Reader

Comment by cy6erlion | Hacker News Reader

If AI can be conscious then every human can voluntarily feel happiness (or any other emotion) anytime anywhere. Everything in a computer depends on symbols, we cannot symbolize happiness with language we can only describe it to others but that is not the same as experiencing it.

View on HN

I think you're thinking about this the wrong way. By the construction of our physical bodies, we generally have no conscious control over many of our internal processes. AIs may not have these limitations, which may result in some very alien forms of consciousness.

For example, you can't decide to see the color blue, but it is very likely that an AI equipped with light sensors could decide to directly provide input to itself as if its sensors were seeing the color blue.

Emotions may (or may not, we don't really understand the working of the brain well enough) work similarly. There may be brain 'circuits' that provide inputs to the part of the brain that creates consciousness that represent emotions. We don't have a way to really 'fake' these internally, but an AI perhaps could. And of course, it is very difficult to imagine how a being that can just choose to be happy in any circumstance would actually think and react to the outside world.

> you can't decide to see the color blue

Once I have seen it I can always visualize the color blue because color can be stored in memory and IMO that is why computers can describe color. But even if I was once happy I can never visualize being happy I just have to experience it and that experience/ substance of happiness cannot be encoded.

Emotions and consciousness are not the same thing.

Emotions are the result of consciousness. So you don't need to be able to build emotions.

Emotions arise in consciousness, but I don't think they're the result of consciousness. You can measure the physiology of an emotion. Thoughts, emotions, sensations, all arise in consciousness, but I don't think any of those are the result of it.

What? Animals are emotional, no? The part of the brain that is responsible of emotions is older and more primitive than the actual thinking part of the brain.

Yes, consciousness is at a higher level, but if we have an entity that can be conscious it should also have the ability to have emotions because we are conscious of emotions.

I don't think that's necessarily true, because I don't think emotions are 'what consciousness is made of'. For example, I'm conscious of color. Imo consciousness is an ability to witness subjectively, in a very specific but non-transmissible way.

Can we have emotions without consciousness?

I think the question is more: can there be a system complex enough to mirror the physiological effects of an emotion, but without a consciousness also experiencing it? Or does the complexity of that system, by definition, cause consciousness to arise? This is sort of getting into the "Philosophical Zombie" problem: https://en.wikipedia.org/wiki/Philosophical_zombie

Thank you for the link, interesting stuff. I think about this all the time, it is interesting to see other viewpoints. I believe that everything we know, feel and are aware of in the universe requires some consciousness as an entry point. I also believe that there are different levels of consciousness, a computer is consciousness to some level but I think there things (emotions etc) that it can never be conscious of simply because everything in its reality boils down to 0's and 1's which I think can never encode happiness otherwise all of us would be happy simply by encoding that state in our minds.

Can you encode the state for '1 + 1 = 2' directly into your mind? How about the state for knowing the solution to the P=NP problem? It is clear that these states exist, but that doesn't mean we have the physical ability to put our brains in them directly. Why would happiness be any different?

Not to mention, there are people who at least claim they can feel at least certain emotions, such as serenity, regardless of the situation that they are in. It may be that with enough study, we ask could actually learn of a way to feel happiness regardless of external context.

IMO yes. Emotions could be thought as input into decision making process. They seem to be just black-box heuristic calculations that one sometimes needs to run inference with proper logic engine to decode correctly (other part of brain).

Tell me difference between sense and emotion? I don't think there's fundamental difference. OK, maybe emotions are just derivatives over senses.

Sense: touch, seeing, smell and so on. One could argue what all senses are, but this is the characterization of it.

Emotion: a process that involves cognitive interpretations of your context (environment + thoughts) and a physiological feeling through the senses (heart rate, tingly sensations, weird feelings in eyes). Check out the James Lange theory. There are better theories, but this has the fundamentals.

I ad libbed this one, I wanted to show that there is a difference. It wasn't my intention to be pinpoint accurate.

Senses are for collecting sense data, emotions are a state of being they are experienced in real-time and can never symbolized or described. You cannot tell someone what happiness is they ave to experience it for themselves. Sensed data can be encoded in language/symbolically for example we can describe music with music notation.

It's fairly easy to voluntarily induce happiness, but it's illegal in many countries.

I see, but what about other emotions (love)

Not sure if love is an emotion, but sure, why not, these likely are just complex activation patterns in the brain.

Is this some kind of new-age philosophy? Show me the part of the brain that doesn't work based on physical properties. If you can't, then we have to agree that the brain is just operating on mathematics.

Computers can do math too. QED: your argument makes no sense.

Reply on news.ycombinator.com