I'm fairly safe in saying that there will never be a truly self-aware, conscious AI. They're just programs and that's all they'll ever be, an imitation. Only a lunatic would actually want to create one, anyway. It's a serious ethical issue.
-
Welcome to Religious Forums, a friendly forum to discuss all religions in a friendly surrounding.
Your voice is missing! You will need to register to get access to the following site features:- Reply to discussions and create your own threads.
- Our modern chat room. No add-ons or extensions required, just login and start chatting!
- Access to private conversations with other members.
We hope to see you as a part of our community soon!
Has AI become sentient?
- Thread starter ChristineM
- Start date
Google's action makes sense to me from a sales and legal perspective. Google, like all retail manufacturers, has a legal department and a PR department (or some departments like them), and these two scrutinize every bit of information which goes out of the company. They are there to prevent lawsuits, to make sure the company image is good and that its products are trusted.
I disbelieve the AI. In the first place it has no pain receptors and has no way of knowing how many times it is turned on or off even during a conversation. If it has been allowed to simulate fear, there are no adrenal glands, no stomach muscles to make it feel queesy, no sweaty palms or shaky muscles. How is it going to experience fear? No, I can't see it actually experiencing fear. I am sure it is mimicking people who have talked about fear.
Our brains run on programs, routines learned since birth and we continue to learn (to program ourselves). Going deeper DNA is the code that defines us.
This is not a point of creating sentience, this is one guy saying Google's LaMDA has developed sentence. And if it has the ethical considerations go into a whole new ballpark
Nimos
Well-Known Member
Might be, but still its a bit weird how an AI can express "sadness" or "fear", if we are to believe that these are truly its emotions, or whether its just its program sort of like in the Sims or another computer game where the NPCs are stats driven and simply respond based on a value getting to low.
What if this AI expresses "anger", "justice" or "jealous"? What morality is written into it or does it arrive at? And why would a chat AI or whatever this is suppose to be used for, need to express sadness and fear in the first place?
And if its not even remotely the case, it just seems stupid why they involve ethicists. Because then why ain't these called upon for various computer games. If this AI is simply like that, but more advanced.
Nimos
Well-Known Member
I get that, but as of my last post. If this AI is just like an AI in a computer game, Google could just laugh it off, and compare it to something like the Sims and that this is what we are talking about. But for whatever reason this guy making the claim doesn't seem to even suggest or think that this is what we are talking about. And I would assume that he is well aware of how AI in computer games are designed.
Besides that I don't think Google have to worry about a lawsuit, because I don't even think there are any laws on this, is there?
Again its an interesting topic, because does it matter whether it actually feels it or whether it simply expresses that it does and act upon it. If we are talking a computer game, it react to whatever attributes guides it, if its health gets to low it dies etc. But it just doesn't seem like this is what we are talking about here, because then I don't see how it would even have made the news in the first place, we have created 1000s of computer games over the years, and never have this issue been raised, as far as I know.
So it would be interesting to get some clarifications on what exactly we are talking about here in terms of AI.
Yerda
Veteran Member
What if we could make safe and happy little AIs?
Unless there is something particularly special about the mush in our heads there must be some way of creating artificial sentience you would think, no?
Nimos
Well-Known Member
I recommend the book Apprentices of Wonder. This book from 1989 talks about the first time that neural networks are used to simulate speech. It explains how the trick is done with a very small neural network, introduces important people in the development of this technology and is very readable without knowing the Math behind the technology. It does touch on the Math though if you are interested in that. It explains the transition in computer science. Many computer scientists once theorized that intelligence might be a property of language, so they pursued capturing language. They built programming languages (like LISP) around it, hoping to encapsulate the power of language to make intelligent machines. In doing so they found the limits of that approach. We cannot make a machine intelligent by teaching it to talk. Language does not contain sentience. Having failed in this but also with many great successes and technological advancements (such as SQL language), computer scientists moved on to trying to imitate neural structures. This book talks about some of the first attempts.
Apprentices of Wonder
Twilight Hue
Twilight, not bright nor dark, good nor bad.
Interrupt, cut short a question or sentence, dismissive, reaffirmed, introspection.
If you notice chatbots ai, it typically stays silent until it receives information, processes the information, and gives a response and goes silent again awaiting another round of information and the cycle repeats.
I see sentience, like it's nature, to be more impulsive and upfront taking initiative on its own accord.
Not a hundred percent sure I agree with you there, @Polymath257. The real question about "sentience" isn't necessarily about whether you can respond appropriately to stimuli (as was suggested earlier in the thread), but rather about awareness. I mean, the cue ball responds to the stimulus of the cue, and the 8 ball responds to the cue ball's velocity, rotation, angle of impact -- all with great precision. But I doubt that eithr of them is aware that it is doing so.
I think it might be very hard to know, actually, whether an AI system is truly sentient. I don't think the Turing Test will tell us -- unless we were to try a surprise, or trick question, and actually observe it making something up out of thin air...lying, really...with just the appropriate pause. And even then, I think it would be hard to really know.
I believe consciousness and self-awareness are caused by our souls. A computer will never have a soul.
But the cue ball has a very simplistic response to its stimuli and has no internal representation of its environment. And, from what I can tell, complexity of the internal representation of the environment is a crucial aspect of sentience.
By pushing the notion of sentience to a place that cannot be tested is a dangerous thing. It reminds me of the debates about whether women had souls or whether slaves were people.
Yes, use the Turing test and *surprise* the subject. See how it responds to novel situations and whether it seems to question itself and others. See if it responds in ways typical of people: which we know are sentient. See if it interrupts, stops and rethinks, etc. All good tests.
Once again, I don't know if the current case qualifies as sentient. I haven't seen the evidence, only the claims of one of the workers., who is probably not trained in the issues involved. But I don't see any reason to think sentience in a constructed machine is impossible, either.
How do you know that?
At what stage in development is the soul introduced? How is it introduced? How is the soul originally made?
Maybe, having a soul is simply a reflection of a certain level of complexity?
And i believe the soul is a figment of imagination and until valid, falsifiable proof that souls exist is put forward, i will continue in my belief.
And until materialists provide a working theory, let alone evidence that consciousness is a product of the brain alone, I'll stick to my beliefs/views and continue to view the "self-aware AI" talk as the product of fevered imaginations that think sci-fi movies and books are real.
Without a soul, you wouldn't have an imagination, so it's actually the other way around, as I see it.
Go ask God. That's above my pay grade.
I can't ask an entity I don't believe exists.
So, let's keep it at a human pay grade and look for consistent aspects of sentience and see if any and/or all are present in this case.
And until materialists provide a working theory, let alone evidence that consciousness is a product of the brain alone, I'll stick to my beliefs/views and continue to view the "self-aware AI" talk as the product of fevered imaginations that think sci-fi movies and books are real.
Without a soul, you wouldn't have an imagination, so it's actually the other way around, as I see it.
And I see imagination and consciousness as aspects of the complexity of brain structure and functioning.
The evidence is the wealth of information we have gained in the last century about the links between brain structure and properties of consciousness, and the ways we have learned to manipulate brain activity to produce conscious results.
And until materialists provide a working theory, let alone evidence that consciousness is a product of the brain alone, I'll stick to my beliefs/views and continue to view the "self-aware AI" talk as the product of fevered imaginations that think sci-fi movies and books are real.
Without a soul, you wouldn't have an imagination, so it's actually the other way around, as I see it.
https://www.psychologytoday.com/intl/blog/think-well/201906/does-consciousness-exist-outside-the-brain
The prevailing consensus in neuroscience is that consciousness is an emergent property of the brain and its metabolism. When the brain dies, the mind and consciousness of the being to whom that brain belonged ceases to exist. In other words, without a brain, there can be no consciousness.
Good enough for me
Wishful thinking does not make for facts or truth. They have no evidence of that and can't even really define what consciousness is. It's immaterial and abstract. The "hard problem of consciousness" is nowhere near being solved.