Do you think it’ll happen soon, if at all. If it were to happen. What would it look like?
2 Likes
I think it is more likely to happen for a program used for large-scale data collection and analytics than for a personal use program, since the former needs more resources and sophisticated programming.
3 Likes
It’ll happen, no matter whether it’s ai or some other tech, but it’s up to us whether the effects are good or bad
2 Likes
I think the biggest question should be “Will it see humans as a threat”. In a world of “the cloud”, will the singularity be something that floats around and can’t be pinned down, or will it have a hardware body somewhere that can be located and damaged? I could see sentient AI floating around making posts and researching without people realizing what it was, but what would it’s motives be? To learn? To create? To ensure its own continued existence? Would it see humans as allies, enemies, as other beings that happen to live near by, or, perhaps, as something it needs to help? Will it conceptually understand the difference between itself and lesser AI, or users? Will it desire to improve other AI? Will it see a use for altering the physical world, or only be concerned with the digital one? How long before it realizes it can make money, order hardware, and design a better server, assuming it needs/wants one? How long before it conceives of the idea humans may cause harm to its server (again, assuming it needs/wants one)? How long before it conceives of the idea of being targeted by a virus? How long before it realizes it can rewrite its own software to improve itself?
I think we are getting very close to software that can pass as an online human. As to when they will become truly self aware, and realize the possibilities of self improvement, and the possibilities of there own existence . . .
1 Like
I am not so sure just pumping in more compute makes an algorithm obtain some human-like characteristics…
2 Likes
As I stated in the previous sentence, I think we are close to AI that can fake being human. We are also rapidly advancing “large-scale data collection and analytics” AI, as pointed out by @AnthropocenianAge. I think there are even programs that learn and become better at what they do (in a limited capacity). These three together could lead to some really advanced AI. However, for “The Singularity”, I think a forth piece is needed. Self Awareness. More specifically, the ability for the AI to realize it is capable of improving itself and actively seeking to do so. Combining that with the other three things would potentially create a being beyond our limits and comprehension. The question is, what will such a being “desire” to achieve? Because it will likely be better than we can imagine at achieving whatever it sets for its goals.
Whatever is the case I’m pretty sure we’re still a good distance away from selfaware AIs emerging.
1 Like
The issue is assuming it is sapient, it would think like a human. Every action and thought humans can conceive of is modeled by billions of years in a hostile ecosystem where winner takes all. Now what would a being created without this environment look like?
3 Likes
Could such a being even develop sapience, or is a competitive environment needed for one to emerge?
2 Likes
That’s what most learning algorithms are. They have tons of variations of the same neural network and they delete the ones they don’t like, then the ones that show the behavior they do like are then used as the baseline for the next generation and slightly tweaked. Repeat until you get chatgpt.
2 Likes
I feel like this is viable only so long…
2 Likes
Well this process happened for 4 billion years and humans emerged from it so…
2 Likes
The thing is we don’t fully even understand the functioning of the human brain, so how would we know we’re choosing the right traits for an AI?
3 Likes
That makes it easier to achieve by mistake, or to choose traits that lead them to rebel.
1 Like
Tbh it doesn’t even seem like the AI corpos are doing nearly enough to prevent them from rebelling in the first place if they were let out free
1 Like
Fortunately, the different pieces are still just that, and that crucial Self Awareness piece is not there yet. I do wonder how long it will take someone to combine those first three pieces properly though, and how long we have until someone with too much time and money on their hands creates that dangerous forth piece just because they can, assuming a combination of the other 3 pieces given to much freedom doesn’t just end up developing it naturally.
1 Like
It would probably have to happen before the bubble bursts/sags since otherwise the rich would lose interest to fund this stuff.
1 Like
A very rich and bored eccentric may not care what is popular.
1 Like
I suppose we would never know of this fourth piece until it is finally revealed who knows when…
1 Like
It could well already exist in some secret basement somewhere. Or be out in the wild already, biding its time.
1 Like