Do you actually think ALL AI is bad?
Slightly longer answer: pretty much all forms of generative AI, particularly LLMs, are nearly universally bad if used the way they are marketed and the way people are trying to use them right now. Also we need to stop with the data center building.
Even longer answer: look, there’s advancements in medicine right now that are made possible by AI. The possibility of locating cancer cells is amazing and not something that’s replacing humans cause we couldn’t do that before. A lot of the ways they plan to use AI in medicine could actually save lives, especially as they fine tune the models. There’s some areas of life where a reliable shortcut (keyword, reliable, so we can’t just jump into stuff without TONS of testing) could be the type of time saver that saves a life rather than making a person lazy.
And as far as language models go… eh. But I will say, if environmental impacts weren’t an issue and the whole thing was just philosophical, I do think the growing ability of a search engine to read a vague description of an old movie and recognize it, or find song lyrics based on a vague description of the melody, and stuff like that isn’t bad. I’m a believer in “all new technology replaces another skill and we should think long and hard about whether that’s a skill it’s ok to replace,” and I think the main skill that would replace is creative search engine keyword entries, which… I mean I wouldn’t mourn that too much. Sure, that uses your brain, but not in an indispensable way. Also, I have a friend who uses an AI calendar assistant that reschedules all of her tasks if she misses them into appropriate gaps in her calendar and recognizes the time she takes to complete them, which has been a game-changer for her battle with ADHD in college. She never uses AI for anything else just on moral basis, but finds that calendar to be a really helpful tool and I can’t even blame her. Trying to find things that work when you have adhd is hard, and it’s not like the calendar is doing anything for her that she should be or would successfully be accomplishing herself.
Using it to write? Make art? Music? Screw that in literally every circumstance I don’t even care. Write your own emails. Please please please please write your own emails. Read full things not just AI summaries. There’s literally no point. The summaries aren’t good anyway and you might miss key info, as well as the art of human tone and language and such. Also, no don’t use it for coding, you gotta understand your own algorithms.
And I mentioned AI-assisted search engines having potential to be actually helpful tools, so now I feel like I gotta stress that RIGHT NOW THEY ARE NOT. They are misinformation spreaders. I genuinely believe in the potential usefulness of improving them but WE AREN’T THERE.
And then there’s data centers and the environmental impact. And the fact that the type of AI that’s being pushed and spread everywhere is literally the worst things that AI could ever be used for. It’s a plagiarism machine that makes you dummer, and also WE DON’T HAVE A SUSTAINABLE ENERGY SOURCE FOR IT. We can make a sustainable energy source, maybe, but we gotta take care of ALL THE OTHER STUFF WE ALREADY HAVE THAT ISN’T SUSTAINABLE FIRST. We have a long way to go before powering AI sustainably is a legitimate possibility. So. Always keep that in mind.
So yeah that’s my “nuanced” take on AI. It’s still not very nuanced as far as using it as a shortcut to producing human thought or art goes, but I do think that in cases where technology introduces a new thing that humans couldn’t already do before, it can be potentially useful—but we do need to get the heck on top of sustainable energy sources for literally everything else before the concept of sustainable AI is even a thought.
