I absolutely cannot take LLMs seriously because of one Computerphile video I watched showing how inefficient they are and why so many massive data centers are being built to run this technology. I've always thought that LLMs are great for finding patterns in data that would either be too time-consuming, or too difficult for humans to notice and categorize. That's what they did with AlphaFold and was probably the biggest accomplishment with LLMs so far, and maybe the only good thing to come out from this technology.
>the vast majority of people will be perfectly content to watch AI films, read AI books, play AI games, etc
The vast majority of people don't have the willpower to fight back. That doesn't mean anything.
>Whether you like it or not...
Shut up. I hate this rhetoric because it shifts the LLM debate into something completely different from technology. Yeah sure, societal changes are beyond any single individual's control, but the argument against generating things with LLMs was never about that.
>It won't cease to exist, but it will become a small niche, significant only for the method of its creation, not for its own value.
What is the value of art? Is art inherently valuable? Is art valuable because it has a labor component, and the product of labor is valuable? Is art valuable because it makes you feel things? You can pull an entire debate just from this topic, but I don't think certain people hire humans to make their art when LLMs exist because of "value" or "method of its creation" (labor).
>I predict that human relationships will be the next victim
>I think romance will be next.
>In the future, you can chat with your own AI girlfriend
None of that is new and are already happening, have you never seen those cases of people having a romantic relationship with ChatGPT or Gemini, after a bad divorce or something? Also go to friend.com, I think you'll like it.
>And then, drying your tears, embrace AI
We'll see when it actually happens. Whatever we have right now is not "AI" in the slightest. Intelligence is not just about pattern recognition, in fact I think most mammals can do that just fine, yet only we humans are regarded as "intelligent", as far as we know. LLM-based chatbots like ChatGPT and Gemini do nothing more than predict with high accuracy the next part of a sentence, which tricks you (the chatbot itself isn't purposefuly "tricking" anyone, it's all you) into thinking it is intelligent, because the only kind of intelligence we know is human intelligence. Humans like to write a lot, and we consider people that can write well to be highly literate, therefore highly intelligent.

This whole thing is a grift dude, a scam. There are billionaires, trillionaires, politicians making huge bucks from this. It's all fake, built on a technology that was never created to do what it is being most used for right now. It's not that deep.