Comment by mxkopy | Hacker News Reader

Comment by mxkopy | Hacker News Reader

Parent Full thread

You can't "cut to the core of deep philosophical questions" when you don't know the basics. For example, this

> we can already comprehensively model some biological neurological systems in their entirety, in silicon.

is just patently wrong.

> How can we confidently state we know ChatGPT doesn't use neurological mechanisms, when we don't yet completely understand our own neurological mechanisms, nor emergent mechanisms that may be at work within ChatGPT?

How can we confidently state that the sun isn't alive when we don't ourselves know what life is? I mean, it's got different layers like a cell. It has a source of energy, a lifecycle and it eventually dies. The sun has been around for much longer than chatGPT so I think we should answer this question first.

> While we routinely "imagine" an objective reality that exists beyond our perception

I'll just say this - if you really want to believe you're a brain floating in space, why don't you live life like it's GTA? Because obviously none of it is real, right?

Look up 'solipsism', this isn't a new or for that matter important philosophical position. Unless you want to write about how much fun whatever video game you're playing is in lieu of reality.

> is just patently wrong.

Not really, here's one I found in seconds, that I hadn't even seen yet - where they specifically remark on the similarities to AI networks:

https://www.science.org/doi/10.1126/science.add9330

> How can we confidently state that the sun isn't alive when we don't ourselves know what life is? I mean, it's got different layers like a cell. It has a source of energy, a lifecycle and it eventually dies.

Well, I've learned from your previous comment you're not agreeing with me, but here's another instance I'd have thought you might be.

These are exactly the kinds of questions we should pose and explore. Are such a massive, complex, long-lasting sequences of nuclear processes "alive" or potentially intelligent? Might they have a detectable consciousness?

What are we but a vast sequence of chemical and electrical processes? In that sense, not much different.

(Science has indeed so far rather arbitrarily defined "alive", and this shows in some of the more controversial categories of "life", such as viruses.)

> I'll just say this - if you really want to believe you're a brain floating in space, why don't you live life like it's GTA?

I've never stated that, I've posed a relevant, important philosophical question I think lies at the centre of the issue of whether or not an AI can be considered to "hallucinate". As was the topic. Your response seems to be "no", and that's fine.

> I've never stated that, I've posed a relevant, important philosophical question

If you think this then go publish a paper and see how the community receives it. They decide what's important and relevant, not you.

Solipsism implies there is only the self. However, it is possible that we all have similar enough brains that have evolved for us to share a false reality together, like Donald Hoffman's recent theories. On that note I do often wonder about the sentience of stars, photons, everything in between. So chatgpt is not much of a stretch and is exciting in that context.

The thing is that words are supposed to be useful. If a word can't tell the difference between photons, stars, and everything in between, then it's not very useful anymore. Calling all of these things 'alive' would just defeat the meaning of the word - we can't resuscitate a rock, we can't talk to photons, feed an atom - all these things that we associate with being alive are just rendered inapplicable.

Reply on news.ycombinator.com