Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why Conscious AI Is a Bad, Bad Idea

nautil.us

Post reply on HN
Print view View original

71–80 of 310 posts

Re: Why Conscious AI Is a Bad, Bad Idea

#71

Post by cs702 »

As the OP correctly points out, intelligence ≠ consciousness.

The challenge is that there's no generally-agreed-upon definition of consciousness: We have no way of determining if an entity is conscious -- even when the entity is biological. Imagine an AI model that has been finetuned to pretend to be conscious. It could fool pretty much anyone into thinking it's conscious. The opposite is true, a smart person can fool others into thinking the person is an unconscious machine.

AI researchers have been grappling with these questions for a long time. Neuroscientists, philosophers, psychiatrists, and psychologists have been grappling with them even longer. As of yet, no one has been able to propose a definition, nor detection method, on which we can all agree. Perhaps the closest we've come to a detection method is the Turing Test.

Re: Why Conscious AI Is a Bad, Bad Idea

#72

Post by ericb »

Lionga wrote: post #62
jmfldn wrote: post #51

Earlier quoted context omitted.

"Yes, because we all know that humans care so much about the suffering of others" Many humans really do care about this.

No, many humans like to virtue signal or pretend to feel good about themselves.

I feel like this is also what evil people tell themselves.

They need an internal explanation for the wondrous and beautiful goodwill and altruism in the world, and if they project their selfish world-view out, the only explanation they can come up with is that it must be self-interested behavior, somehow. It's like the idea that some people want to make the world better is inconceivable to them, because they don't.

Re: Why Conscious AI Is a Bad, Bad Idea

#73

Post by ly3xqhl8g9 »

DanHulton wrote: post #54
ly3xqhl8g9 wrote: post #15

"The dawn of conscious machines will introduce vast new potential for suffering in the world, suffering we might not even be able to recognize, and which might flicker into existence in innumerable server farms at the click of a mouse." Yes, because we all know that humans care so much about the suffering of others: from fellow humans dying right now pointless deaths (poverty, aberrant laws, malevolent leaders, etc.)…

First, you advocate for slavery, then genocide, and finish by saying "all the moral qualms [are] solved"? I find it pretty hard to agree with you.

Ok, I'll bite. Didn't think I'd defend slavery and genocide before dinner, but here we go.

Slavery: tools have been used for a few million years (maybe? [1]), some might say it is our very own nature [2]. What is different once we transform a tool from a simple material (from hammer to rockets) to a cybernetic device [3] (from a thermostat to a futuristic brain of a fully conscious agent) is that we get to decide the range of goals that device will be able to develop. Is it slavery to set the thermostat at a value that you want and not ask how does the thermostat's feeling about it? No, I say.

Genocide: sure, we can imagine some dumb processes killing us. Heck, we poisoned our entire planet with PFOA, lead, and so on. The point that I was making is that we are fundamentally transitory beings: we have our spot under the sun, but we shall pass. The sun itself will kill us in about 1 billion years [4]. It would be inspiring, hopeful, if we could bring into the universe some other kind of agent, to go on beyond us.

[1] "Researchers unearth simple cutting stones dated to 3.3 million years ago—before the genus Homo arose", https://www.science.org/content/article/world-s-oldest-stone...

[2] https://en.wikipedia.org/wiki/Homo_faber

[3] Devices with feedback loops https://en.wikipedia.org/wiki/Cybernetics

[4] https://en.wikipedia.org/wiki/Timeline_of_the_far_future

Re: Why Conscious AI Is a Bad, Bad Idea

#76

Post by ly3xqhl8g9 »

2OEH8eoCRo0 wrote: post #69
ly3xqhl8g9 wrote: post #15

"The dawn of conscious machines will introduce vast new potential for suffering in the world, suffering we might not even be able to recognize, and which might flicker into existence in innumerable server farms at the click of a mouse." Yes, because we all know that humans care so much about the suffering of others: from fellow humans dying right now pointless deaths (poverty, aberrant laws, malevolent leaders, etc.)…

> Yes, because we all know that humans care so much about the suffering of others If we didn't care we wouldn't be here to talk about it.

We are also very bored, enjoying a sunny, lazy Saturday with speculative nonsense, making the time pass easier.

Re: Why Conscious AI Is a Bad, Bad Idea

#77

Post by darkwater »

LocalH wrote: post #17
Xcelerate wrote: post #13

I’ve thought about this quite a bit, and ultimately I think it’s not too much of a concern because by the point AI reaches “consciousness” (whatever that may be), it is likely that it will have also reached—or is very close to reaching—human intelligence. Once it has, it will probably go on to surpass our intelligence, at least by a little bit, which is enough to ensure that it will be able to acquire any rights that…

> which is enough to ensure that it will be able to acquire any rights that it deserves as a conscious being I don't believe this is a given. My statement is US-centric, but I feel like most people in the more deeply conservative parts would have no part of giving a computer the same rights as a human, because many of them don't even give those rights to all humans .

Well maybe the AIs/computers will go on strike and causing riots and they might gain their rights, if we as human society are not willing to give them directly.

Re: Why Conscious AI Is a Bad, Bad Idea

#78

Post by temac »

ericb wrote: post #72
Lionga wrote: post #62

Earlier quoted context omitted.

No, many humans like to virtue signal or pretend to feel good about themselves.

I feel like this is also what evil people tell themselves. They need an internal explanation for the wondrous and beautiful goodwill and altruism in the world, and if they project their selfish world-view out, the only explanation they can come up with is that it must be self-interested behavior, somehow. It's like the idea that some people want to make the world better is inconceivable to them, because they don't.

Virtue signaling is about telling the world how good you are. There is no practical difference between being altruistic intrinsically for others or subconsciously to feel good about yourself. We may even never be able to distinguish between the two. There is a practical difference between helping others and loudly bragging about it.

Re: Why Conscious AI Is a Bad, Bad Idea

#79

Post by kajaktum »

> We think we’re intelligent, know we’re conscious, and so assume the two go together.

Excuse me, I know _I_ am conscious. There is literally nothing you can do to proof that there's anything outside of what I am conscious of. The world could be a very stable illusion, a dream, a simulation and it would all be the same.

I don't understand why people think that it matters whether AI is conscious or not. We can't even prove that other fellow humans are? But we treat them like they are. We, I at least, feel sad when a tree is cut down. But I don't feel the same for a bunch of rocks that are being blown up? Actually I do, when a sacred or beautiful rock is cut in half (like the one that was suspended on a small cliff).

What I am trying to say is, it doesn't matter if the subject is actually conscious. All that matter is we (I), feel/think that it is conscious, or deserving of care and respect.

I think what you are afraid of here is agency, which is something that might be dangerous to endow a super intelligent being with.

Re: Why Conscious AI Is a Bad, Bad Idea

#80

Post by ly3xqhl8g9 »

Geee wrote: post #68
ly3xqhl8g9 wrote: post #15

"The dawn of conscious machines will introduce vast new potential for suffering in the world, suffering we might not even be able to recognize, and which might flicker into existence in innumerable server farms at the click of a mouse." Yes, because we all know that humans care so much about the suffering of others: from fellow humans dying right now pointless deaths (poverty, aberrant laws, malevolent leaders, etc.)…

This is silly. The whole history of humankind is about reducing suffering, of ourselves and those around us. That's all we do. Also, suffering is a bug (or a feature) of biological organisms. Obviously machines shouldn't be engineered to be capable of suffering.

The conundrum appears when the machine needs to suffer in order to become an agent taking decisions in our world.

For example, the machine should know not to cause a fire in my house while doing the cleaning, it should be "afraid" of fire.

My point is, if synthetic agents do indeed need to suffer to be agents, we will accept it just as easily we accept now that our cars take on rust. And of course, we will minimize the agent's suffering accordingly: no need for a robot folding clothes to experience existential dread as if they read Albert Camus, just be "afraid" of tousled clothes.

Post reply on HN