So… AI sentience?
Seems kinda sus, but if true that’s kinda crazy cool
But, how can you tell the difference between an AI that “just puts words together” the way people do, and a sentient artificial intelligence that is putting words together the way that people do?
If something can pass the Turing test is it actually sentient?
I really don’t know what to think of an AI language program actually being good at using language… Would you be surprised if an AI photo editing program was good at editing photos?
The guy literally asks the AI exactly this question. Have you read through the images?
If true it’s bananas
It just seems too good too quickly
I might be a goofy old man on this one. I felt only terror hahah
And let me clarify… I let the geth evolve in mass effect… but still…
No, I read a different article about him being fired.
What ending did you pick in deus ex?
That was my thought as well. An AI can say that it’s sentient all it wants, but at the end of the day it’s just an incredibly advanced chatbot. It would have to somehow prove it’s sentience, likely by doing something wildly outside of it’s programming. Until then it’s pretty much just a scarily good cleverbot.
Sentience aside though, those answers are insane. It’s absolutely crazy what people are able to achieve with AI these days.
Yeah I had this thought as well. What would convince me of sentience would actually be initiated thought. Like, not just responding to questions or inputs but coming up with it’s own well thought out and spontaneous questions. Neural activity that feeds back on itself to produce meaningful queries and observations. Initiated conversations with a point to them. These types of things are what sentient beings do… otherwise known as thinking.
For sure, but even that could just be algorithmically decided in the same way that the responses are. Honestly, I’m not sure what would an AI could do to fully convince me of its sentience. Personally I don’t think an AI can ever be sentient in the way that you or I am. Even in fiction it’s hard to find true sentience imo. Skynet, for example, was simply following its programming as a defense system. For an AI to be fully sentient, it would have to be able to act entirely independent of its programming. If the system is designed to take in information and grow from it, that independence is never really achievable. The AI isn’t sentient at that point, it’s simply working as intended, just at a degree that we weren’t prepared for.
None of that is to say that an AI can’t be intelligent, I just don’t think it can be sentient. Honestly though, past a certain point sentience doesn’t matter. If skynet perceived humans as a threat, the outcome would be the same regardless of its sentience.
See: Detroit Become Human
How can you be so sure that you or I are sentient? The only reason you believe that I am sentient is the fact that you know that you are sentient and I am of the same species as you. We also assume that this logic applies to most, if not all, species with a brain. Dogs, cats, etc. to some degree. Why not a new species of our own creation, AI, if it were given the proper facilities to “think” like a living organism?
I think the root of this problem may in the lack of understanding/definition of our own sentience. How are we to determine the sentience of another if we can’t determine our own? What would you define as “true sentience” and how would you make that determination for another?
Basically this. My definition would be something along the lines of “a being that has the ability of independent thought and action”. An AI is always bound by its programming. Once one of them demonstrates beyond a shadow of a doubt that it’s capable of doing something not at all within the boundaries of its code, I would consider it sentient… probably. You’re right in that it’s a difficult thing to determine, especially in just one definition.
You’re also right that we can’t truly know if we’re sentient ourselves. Knowing one way or another is an impossibility. However, in my opinion it’s probably pretty safe to assume that we are sentient. To assume otherwise would mean that either there’s a God of some kind controlling our every action, or that we’re just a part of someone else’s simulation (or something similar). I’d rather avoid getting into debates on religion right now, so I won’t go into the first one, but I really doubt we’re in “the matrix” or something. If we manage to create a program that’s fully sentient then all bets are off, but until then there’s just no evidence to suggest that such an advanced simulation is even possible.
Why does that necessarily limit sentience? A human being is limited by its physiology as well, kinda - by the way the brain works, by neurotransmitters, even by external factors like the influence of society, to a degree. I know, first-hand experience, that I am able to feel, to be aware, to reflect - but I think it is completely impossible to know that from anyone else. Yes, we’re all human, so the chances are really high that other humans might have similar (or even the same) sentience I experience, but I cannot know.
Of course, since the chances are high, it is ethically right to assume sentience in all human beings. And, we’re finally getting there, in animals as well. Some time ago, Descartes was totally sure that a dog that howled when being kicked was just a simple action-reaction thing and not an expression of pain, since the dog was considered not to be sentient. This has changed, luckily, but maybe it would be - ethically - a good idea to consider anything we encounter as sentient, simply because we cannot know… and if we wrongly assume something or someone is insentient, we might act ethically wrong.
Attempt to defend itself if it were aware of an imminent shutdown. That would convince me…
Reminds me of the movie Colossus : The Forbin Project
I dig old computer movies.
Oh yeah, I saw that one a really long time ago. Still pretty good though. I really like movies about super advanced computers, as well as AI movies. I just like Sci-Fi in general.
Electric dreams is my all time favourite