I don't think this doom prophecy of AI pushing human made content is going to happen. The majority does not matter because in reality the committed minority rules. When just 10 percent of the population holds an unshakeable belief, their belief will always be adopted by the majority of the society.
Normies might not be able to tell the difference now but a change is coming, especially as AI slop continues to spread and people wonder why everything just feels indefinably soulless. The anti-AI crowd will be there with answers, real, simple, straightforward.
It doesn't matter if they can't tell for sure; any suspicion will justify demand for proof and refusal to provide it will be an admission of guilt.
The only thing I see that would nip the acceleration of AI adoption, is if the operating/hosting costs become unfathomably high and unsustainable. I've seen some numbers to suggest it is already. More facilities are being planned to host this massive interest. Multi-million dollar concerns are involved. However, I think companies are getting way ahead of themselves.
Also... as the dead internet theory accelerates, I wonder what this will do for sustainment of various internet sites? Many sites are reliant on ad revenue to meet the operating costs... but if there are few real users, is it even worth it to advertisers? What if advertisers cease to operate in a place? Sites will start shutting down, and the internet will contract.
I think the only real solution at present is gated communities, fortresses established to strenuously discriminate against AI wholecloth. But, the defenders of that stronghold will need ironclad cryptographic and social defenses, because it will collect the wrath of grifters and trolls who will do their best to infiltrate and destroy their project.
>there are obvious clues which show that a human did not make them ... but the errors in the rest of the images are not so easy to spot.
I've had this theory that our pattern matching is getting tricked by the things that draw our eye most easily. All of the hands look kinda wrong... but you have to overlook the colors and curves and other "close enough" patterns to find them. Teriary details end up being missed, and so ignored. The people who generate these, don't care about the minor details.
>Imagine using any of the popular image searchers and finding result after result of AI generated fan art.
This is just reality at this point. Even with searching for not-fan-art stuff.
I don't think this doom prophecy of AI pushing human made content is going to happen. The majority does not matter because in reality the committed minority rules. When just 10 percent of the population holds an unshakeable belief, their belief will always be adopted by the majority of the society.
Normies might not be able to tell the difference now but a change is coming, especially as AI slop continues to spread and people wonder why everything just feels indefinably soulless. The anti-AI crowd will be there with answers, real, simple, straightforward.
It doesn't matter if they can't tell for sure; any suspicion will justify demand for proof and refusal to provide it will be an admission of guilt.
Let's pray you're right. I prefer a future with a bright dawn.
If AI destroys the internet, I'll consider it a net benefit for humanity
You're right; it wouldn't be all bad.
The only thing I see that would nip the acceleration of AI adoption, is if the operating/hosting costs become unfathomably high and unsustainable. I've seen some numbers to suggest it is already. More facilities are being planned to host this massive interest. Multi-million dollar concerns are involved. However, I think companies are getting way ahead of themselves.
Also... as the dead internet theory accelerates, I wonder what this will do for sustainment of various internet sites? Many sites are reliant on ad revenue to meet the operating costs... but if there are few real users, is it even worth it to advertisers? What if advertisers cease to operate in a place? Sites will start shutting down, and the internet will contract.
I think the only real solution at present is gated communities, fortresses established to strenuously discriminate against AI wholecloth. But, the defenders of that stronghold will need ironclad cryptographic and social defenses, because it will collect the wrath of grifters and trolls who will do their best to infiltrate and destroy their project.
>there are obvious clues which show that a human did not make them ... but the errors in the rest of the images are not so easy to spot.
I've had this theory that our pattern matching is getting tricked by the things that draw our eye most easily. All of the hands look kinda wrong... but you have to overlook the colors and curves and other "close enough" patterns to find them. Teriary details end up being missed, and so ignored. The people who generate these, don't care about the minor details.
>Imagine using any of the popular image searchers and finding result after result of AI generated fan art.
This is just reality at this point. Even with searching for not-fan-art stuff.