I don't think AI output on some factual topic is comparable to distinct things written with IDEs.
On a given topic, I have always found that AI comes to the point of the average talking points of that topic and you really can't cleverly get more out of it because that's all that it "knows" (ie, push back gets either variations on a theme or hallucinations). And this is logical a given method is "average expected reply".
Genericness is overwhelmingly a product of RLHF rather than an innate property of LLMs. A lot of manual fine-tuning has gone into ChatGPT and Gemini to make them capable of churning out homework and marketing blogs without ever saying anything offensive.
If you make requests to the Sonnet 3.5 or DeepSeek-R1 APIs and turn up the temperature a little bit, you will get radically more interesting outputs.
Isn’t that still pulling from the same distribution with a larger standard deviation? I think the problem here is that it only covers a small part of the search space. I think the problem here is that generators are not using novel distributions. They’re still sampling from the same population (existing written works).
On a given topic, I have always found that AI comes to the point of the average talking points of that topic and you really can't cleverly get more out of it because that's all that it "knows" (ie, push back gets either variations on a theme or hallucinations). And this is logical a given method is "average expected reply".