The death of Stackoverflow is one of these events where the site has been completely killed by AI and yet its contents is completely necessary for AI to know about solving programming problems. Its death will mark the end of AIs ability to learn how to solve programming issues. Its cannibalizing itself in the process, as it destroys its sources it destroys its own ability to learn.
It’s not just that, it’s shitting where it eats. People are using it to fill the internet with disinformation, then it trains itself on it’s own disinformation, and breeds even worse disinformation. This is why AI can never be smarter than it was in 2021.
On top of that, due to the indiscriminate DDOSing of the entire internet by AI bots, websites have been blocking any web crawlers that are not Google, which just contributes to their monopoly.
This is why AI can never be smarter than it was in 2021.
You’re completely wrong. First of all, datasets are getting bigger and objectively better. Secondly, technologies and methods of model training become dramatically more complex. Yes, AI content can create echo for itself, but it’s a solvable issue.
Anyone who disagrees with this fact should do a reality check and learn a little more about the current AI state. In 2021 we didn’t even have reasoning in models.
What it have to do with OpenAI? It’s not what I’m talking about.
Nonetheless, it’s kind of fun to see so much downvotes on my comment. Some people are so blinded by their hate (well deserved, by the way) that they will deny something that obvious.
Our key finding is that by injecting information through an external synthetic data verifier, whether a human or a better model, synthetic retraining will not cause model collapse.
Yeah if you have a source of truth then your model is basically getting trained on that.
My point was that having a verifier means your not really training a model on another model’s data, it’s basically as if you get new raw data from a non AI source
Our key finding is that by injecting information through an external synthetic data verifier, whether a human or a better model, synthetic retraining will not cause model collapse.
Lol, so to make a great model, they just need to have an even better one available first or a human who can verify every single thing it ingests.
This assumes everything is valid on the external. If one slop cluster feeds off another - a slopveyor? - then there is nothing external for the validation hall-monitor to compare against. They’re trusting another model’s output as if it were gospel.
It is evident that it has intelligence, it outputs intelligent responses usually adequate to its input, even if it’s badly phrased. What it doesn’t have is sentience, conscience, and a learning loop.
To anyone who interacts with it? Would you deny that a program automate mental labor in the same way that a sawmill automates manual labor? Isn’t that some degree of intelligence?
Now, we have very imperfect LLMs who nevertheless can be instructed not with program code, but actual natural language, and they react accordingly. Isn’t that also intelligence? Computers that understood natural language was the realm of science fiction just five years ago.
I get it that people hate LLMs, both because how idiots use them, and how corporations push them everywhere; but not recognizing the intelligence in those programs is naive at best.
The death of Stackoverflow is one of these events where the site has been completely killed by AI and yet its contents is completely necessary for AI to know about solving programming problems. Its death will mark the end of AIs ability to learn how to solve programming issues. Its cannibalizing itself in the process, as it destroys its sources it destroys its own ability to learn.
It’s not just that, it’s shitting where it eats. People are using it to fill the internet with disinformation, then it trains itself on it’s own disinformation, and breeds even worse disinformation. This is why AI can never be smarter than it was in 2021.
On top of that, due to the indiscriminate DDOSing of the entire internet by AI bots, websites have been blocking any web crawlers that are not Google, which just contributes to their monopoly.
You’re completely wrong. First of all, datasets are getting bigger and objectively better. Secondly, technologies and methods of model training become dramatically more complex. Yes, AI content can create echo for itself, but it’s a solvable issue.
Anyone who disagrees with this fact should do a reality check and learn a little more about the current AI state. In 2021 we didn’t even have reasoning in models.
I see someone’s been suckling the OpenAI teet.
What it have to do with OpenAI? It’s not what I’m talking about.
Nonetheless, it’s kind of fun to see so much downvotes on my comment. Some people are so blinded by their hate (well deserved, by the way) that they will deny something that obvious.
You are the only one being blinded by AI marketing and hype.
Ok, lol.
Model collapse isn’t a thing anymore. https://arxiv.org/html/2510.16657v1
Yeah if you have a source of truth then your model is basically getting trained on that.
It’s like already having the answer
The point is that it only needs to comprise a very small part of the model.
My point was that having a verifier means your not really training a model on another model’s data, it’s basically as if you get new raw data from a non AI source
Lol, so to make a great model, they just need to have an even better one available first or a human who can verify every single thing it ingests.
Hmm, call me skeptical on this claim.
This assumes everything is valid on the external. If one slop cluster feeds off another - a slopveyor? - then there is nothing external for the validation hall-monitor to compare against. They’re trusting another model’s output as if it were gospel.
LOL OK
I’m pretty sure AI is objectively smarter today than it was 5 years ago.
Since LLMs literally can’t learn, no. They’re just increasingly tweaked to seem even more convincing.
How can something with no intelligence be smarter?
It is evident that it has intelligence, it outputs intelligent responses usually adequate to its input, even if it’s badly phrased. What it doesn’t have is sentience, conscience, and a learning loop.
Evident to whom exactly?
To anyone who interacts with it? Would you deny that a program automate mental labor in the same way that a sawmill automates manual labor? Isn’t that some degree of intelligence?
Now, we have very imperfect LLMs who nevertheless can be instructed not with program code, but actual natural language, and they react accordingly. Isn’t that also intelligence? Computers that understood natural language was the realm of science fiction just five years ago.
I get it that people hate LLMs, both because how idiots use them, and how corporations push them everywhere; but not recognizing the intelligence in those programs is naive at best.
No, none of that is intelligence. By any definition.
I have deep concerns as to the competence of anyone who interacts with these LLMs and hallucinates that they are intelligent.
There’s better integration with all sorts of other sources of truth beyond the LLM training, which makes it seem smarter.
Objectively + smarter, huh.
It is not alive and cannot really think, so I doubt it’s smarter. It likely contains a bit more knowledge and a better interconnected network for it
Whether said knowledge is of better quality than 5 years ago is up for debate though.
Actually it just appears smarter because people are objectively dumber than they were 5 years ago. “AI” is actually stagnate.
And yet, they have not created the AI that could do without it. Any day now, promise!
There would be a wisdom in AI companies fighting to NOT have their products in use on websites that produce the best handle content for them to eat.