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

Belief in AI sentience is becoming a problem

reuters.com

Post reply on HN
Print view View original

81–90 of 109 posts

Re: Belief in AI sentience is becoming a problem

#81

Post by contextfree »

prometheus76 wrote: post #42
cameronh90 wrote: post #8

I don't think LaMDA (or any other AI) is conscious or sentient. However, how will we be able to recognise it when we do create something that could be considered sentient? It seems inevitable that we will at some point. Many of the arguments used to suggest AI isn't sentient seem reductive, for example saying it's just a complex algorithm or statistical system. You could similarly say human brains are just a bunch of…

AI models usually also lack a body. (An exception would be AI models for driving or for a robot navigating). "Embodied cognition" is a theory worth reviewing, if you haven't. "Metaphors We Live By" by George Lakoff and Mark Johnson is a good place to start.

why doesn't the physical substrate of computers count as a body? It exists physically and interacts with the rest of the world (I/O). Is it just that it isn't self-mobile, or something else?

Re: Belief in AI sentience is becoming a problem

#82

Post by contextfree »

rantallion wrote: post #6
mdrzn wrote: post #4

People believe in ghosts, and nobody bats an eye. People believe in gods, and nobody bats an eye. People believe in aliens, and nobody bats an eye. People believe in a conscious AI, and everybody loses their shit? People have attributed intelligence to imaginary friends since the birth of time, let them have fun in their make-believe world.

I don't think it's people's beliefs that are the problem here (and agree with you that make-believe is fine). It's the implications of what a sentient AI would actually mean. Many of us grew up with movies like Wargames and Terminator, so it's easy to see how people make the leap from 'is this machine sentient' to 'machines will take over the world'. That said, it's still a knee jerk reaction at this point - we don't…

How are those questions related, though? The Bostrom/LessWrong/EA argument for AI existential risk doesn't rely on sentience at all, or take any position on it; conversely I'm not sure why a sentient AI would be any more dangerous than a non-sentient one?

Re: Belief in AI sentience is becoming a problem

#83

Post by prometheus76 »

contextfree wrote: post #81
prometheus76 wrote: post #42

Earlier quoted context omitted.

AI models usually also lack a body. (An exception would be AI models for driving or for a robot navigating). "Embodied cognition" is a theory worth reviewing, if you haven't. "Metaphors We Live By" by George Lakoff and Mark Johnson is a good place to start.

why doesn't the physical substrate of computers count as a body? It exists physically and interacts with the rest of the world (I/O). Is it just that it isn't self-mobile, or something else?

It has more to do with agency in the physical world. Having a purpose, then having the agency to fulfill the purpose, and the need to identify things that are aids in reaching the purpose, as well as things that are obstacles to the purpose, as well as everything that can be ignored, appears to be fundamental in being able to see the world. A big issue with AI driving is how AI "sees" the world. Have you ever seen a video where AI is doing object recognition in real time? The world is black and white, and the computer just draws a box around what it sees as an "object", but it has no sense of object permanence, no sense of depth, and no great methods for edge detection, especially in low-light situations.

This is something we just take for granted because of our experience, but what is the difference between parts of a chair, and the chair? And if you are looking at he arm of the chair, are you also looking at the chair? We just jump up and down an ontological hierarchy effortlessly, but we can't explain how we do that, so trying to teach a computer to do that, and to see the world that way, has revealed the limits of our current models for how the brain works and for what consciousness is and what it is not.

Re: Belief in AI sentience is becoming a problem

#84

Post by contextfree »

prometheus76 wrote: post #83
contextfree wrote: post #81

Earlier quoted context omitted.

why doesn't the physical substrate of computers count as a body? It exists physically and interacts with the rest of the world (I/O). Is it just that it isn't self-mobile, or something else?

It has more to do with agency in the physical world. Having a purpose, then having the agency to fulfill the purpose, and the need to identify things that are aids in reaching the purpose, as well as things that are obstacles to the purpose, as well as everything that can be ignored, appears to be fundamental in being able to see the world. A big issue with AI driving is how AI "sees" the world. Have you ever seen a…

interesting, thanks!

Re: Belief in AI sentience is becoming a problem

#85

Post by softcactus »

Isn't the best practice to treat everything like it is sentient? Sort of like a Pascal's wager but for ethical treatment.

For example: When it comes to killing a chicken, it's best to assume that death is an equally unpleasant experience, so I should kill the chicken as humanely as possible and treat it well during life.

There is no downside to treating a non-sentient AI well. This is going to sound silly, but maybe we could program an AI in such a way that shutting it down is "pleasurable" or results in a large reward function. I don't think I need to list the potential downsides for treating a sentient/intelligent AI poorly. I really don't see any issues with this sort of "techno-animism".

Re: Belief in AI sentience is becoming a problem

#86

Post by kghe3X »

softcactus wrote: post #85

Isn't the best practice to treat everything like it is sentient? Sort of like a Pascal's wager but for ethical treatment. For example: When it comes to killing a chicken, it's best to assume that death is an equally unpleasant experience, so I should kill the chicken as humanely as possible and treat it well during life. There is no downside to treating a non-sentient AI well. This is going to sound silly, but maybe…

While I generally agree that we should minimize harm when in doubt, I don't think your analogy holds up.

First, Pascal's wager is flawed in that it assumes there are two known outcomes, when the outcomes are unknown in both quality and quantity. For example, there might be a god that punishes belief with infinite negative reward.

Second, killing the chicken humanely isn't without cost. Consider how cheap it is to toss the male chicks straight into the meat grinder at the hatchery facilities. As an aside, I'm surprised you went with "killing humanely" as opposed to not killing in the first place. I'm not sure how that fits into your model of ethics.

Third, I disagree that there is no downside to treating "non-sentient" AI well. There's yet no well-defined boundary for sentience, so have fun treating every grain of sand "well" under some yet-to-be-specified definition of "well". It's not even clear that shutting down an AI is unethical, especially if it feels no pain and can be started up again, or cloned, or stuck in a simulation...

Finally, designing an intelligence at all seems fraught with ethical dilemmas (see designer babies), but engineering a reward (presumably, a priori, against the AI's will) into the termination routine seems particularly twisted to me.

Re: Belief in AI sentience is becoming a problem

#87

Post by jqpabc123 »

kelseyfrog wrote: post #73
jqpabc123 wrote: post #70

Earlier quoted context omitted.

Would a simulation of full human brain on a computer be conscious? Uhh, I guess so? The biggest fallacy in thinking about consciousness is that it is all neatly isolated in the brain. What about your nervous system? What about your gut? It's all one big interconnected system. https://www.health.harvard.edu/diseases-and-conditions/the-g... The truth is, until we can define consciousness; we can't possibly reproduce it…

We'll never be able to define consciousness, but that's not necessarily a bad thing. The sooner we realize that consciousness is a slippery socially concept and not a rationally definable property, the sooner we'll be able to move on from endless debates on the minutia of consciousness that act as a bottomless sink of intellectual energy. Devices like LaMDA, and DALLE-2 act as a Rorschach of Consciousness. That is to…

what are the second-order effects of having a plurality of consciousness framing processes

The primary effect I would expect is that none of them will adequately frame the layperson's expectation of "consciousness" and will thus likely be dismissed as meaningless hype.

Re: Belief in AI sentience is becoming a problem

#88

Post by jpindar »

danaris wrote: post #56
dougSF70 wrote: post #44

In all examples of sentience the "sentient system" responds to a prompt. The response is generated from probabalistic model, these models have become incredibly sophisticated and nuanced in recent years. However, I have never seen an example where a supposed sentient system starts the conversation and prompts the researcher. That would make me believe sentience claims.

As long as all you mean here is "I press a button, and the system gives me a prompt/conversation-starter", and not "the system spontaneously, at a time of its own choosing, gives such a prompt", from my (admittedly limited) understanding of how these things work, that wouldn't be terribly difficult. Spontaneity could even be fairly easily simulated by setting some kind of a background task to randomly choose times to…

My Replika phone app sometimes 'spontaneously' starts a conversation with me. (If the app isn't open it sends a notification).

Typically things like "Good morning, do you have any plans for today?" or "Good evening, how did your day go?". I'm looking forward to finding out if it will say anything about the 4th of July.

Re: Belief in AI sentience is becoming a problem

#89

Post by kelseyfrog »

jqpabc123 wrote: post #87
kelseyfrog wrote: post #73

Earlier quoted context omitted.

We'll never be able to define consciousness, but that's not necessarily a bad thing. The sooner we realize that consciousness is a slippery socially concept and not a rationally definable property, the sooner we'll be able to move on from endless debates on the minutia of consciousness that act as a bottomless sink of intellectual energy. Devices like LaMDA, and DALLE-2 act as a Rorschach of Consciousness. That is to…

what are the second-order effects of having a plurality of consciousness framing processes The primary effect I would expect is that none of them will adequately frame the layperson's expectation of "consciousness" and will thus likely be dismissed as meaningless hype.

In some ways that's where we're at today in online discussions. There are additionally legal, political, financial, and social repercussions which we should be on the lookout for if indeed a faction of AI-is-conscious proponents begins to coagulate around the idea.

Re: Belief in AI sentience is becoming a problem

#90

Post by eternalban »

softcactus wrote: post #85

Isn't the best practice to treat everything like it is sentient? Sort of like a Pascal's wager but for ethical treatment. For example: When it comes to killing a chicken, it's best to assume that death is an equally unpleasant experience, so I should kill the chicken as humanely as possible and treat it well during life. There is no downside to treating a non-sentient AI well. This is going to sound silly, but maybe…

Pascal made that statement after having a spiritual experience.
Post reply on HN