Raw LLM Responses

AI training is no different to any previous kind of statistical analysis. It also happens to be no different to how our brains perform such analysis. And yet these people who oppose generative AI, did not have a problem with other humans learning from the data they put out, nor did they complain about prior revenue generating statistical analysis tasks, when we used the same data to produce marketing information and such. It only became a "copyright" issue the moment the data started being used to produce new data which they feel is in competition with their own. They feel like this new technique has a competitive advantage, because they didn't mind humans that are just as slow as they are, learning from the art they see. Funnily enough I have not seen actual good artists that feel threatened by AI. Most of the people losing their shit on twitter and reddit, if you take a look at their portfolios, they do anime and furry NSFW content, I swear to god. I like to browse artstation and download digital art that I vibe with, and even there I've noticed an inverse correlation between the quality of the art someone posts and the likelihood they've posted that No-AI image. A truly good artist feels secure in the value he brings to the world. It's these people producing "art" of the same quality as something a 9 year old could produce, that feel deeply threatened and violated by AI adjusting a couple synaptic weights by some ~0.00001 because of their 100kb image out of a 90tb dataset to ultimately produce a 9gb parameter checkpoint. Are we going to see world famous art pieces suddenly drop in values? Will the Mona Lisa become worthless because people can generate similar images 5 times per minute? No. That's absurd. Human art was never about what image or sound you can produce. It was about the meaning, the experience, getting to step into someone else's shoes, being able to feel what another human once felt even a thousand years apart. But then, that doesn't apply when you make my little pony tributes. Neurons seeing and learning concepts is suddenly problematic when those neurons can output directly to a screen. We wouldn't need these image generating AIs if we could just imagine things and render them to a screen. Instead we need to imagine concepts and type the prompt, and hope the AI is able to imagine the thing we imagined on our behalf because its neurons can output digital information. Otherwise, the process is no different. So what they truly have a problem with is not having to pickup a pencil and draw the art. Let's hope the neuralink users never start being able to output their mental images to a display or we'll have a genocide of the paralyzed. These people are unhinged. Live and let live.