Ask Lemmy

It's funny to see Godzilla in weird contexts.

No, I don't think that's a particularly good reason for it all, either.

I was really psyched about AI when it first hit my news feed. Now I'm less than impressed. Most generalist AI platforms get things wrong constantly. Having an LLM trained on specific things, like math or science or maybe law, I could see being useful.

We're at the "AI everything" phase instead of the "AI what makes sense" phase.

I find ChatGPT useful in getting my server to work (since I'm pretty new with Linux)

Other than that, I check in on how local image models are doing around once every couple of months. I would say you can achieve some cool stuff with it, but not really any unusual stuff.

It stimulates my brain, and I enjoy the randomness of it all. It's like how in nature things can be perfectly imperfect - random and still beautiful - unintentional and still emotion-inducing. Sure, I see the ethical issues with how an AI is trained and how capitalism cares more about profit than people leading to job loss or exploitation; however, those are separate issues in my mind, and I can still find joy in the random output of an AI. I could easily tunnel on the bad parts of AI and what's happening as the world devours a new technology, but I still see benefits it can bring in the medical research and engineering fields.

I use AI every day. I think it's an amazing tool. It helps me with work, with video games, with general information, with my dog, and with a whole lot of other things. Obviously verify the claims if it's an important matter, but it'll still save you a lot of time. Prompting AI with useful queries is a skill set that everyone should be developing right now. Like it or not, AI is here and it's going to impact everyone.

I've found AI to be incredibly helpful for me.

It's an overly broad term, and the "hype" use-cases dominate the discussion in a way that lacks vision. I'm using machine learning to optimize hardware accelerated processing for particle physics. So, ya, it's not all slop. And what is, may very well evolve.

I’m not impressed with the LLMs. They do make great synonym generators.

Stable diffusion and other image diffusers are genuinely amazing. And I’m not talking about asking copilot to make Fortnite shrek. There are incredibly complex ways in which you can fine tune to tell it how to shape and refine the image. It has and is going to continue to revolutionize graphical art. And once the math shrinks down it’s going to be everywhere.

I like messing with the locally hosted AI available. We have a locally hosted LLM trained on our command media at work that is occasionally useful. I avoid it otherwise if I didn't set it up myself or know who did.

Only one I ever use is the meta AI built into messenger because my friends and I can have it make silly and often extremely cursed pictures that make us laugh

I used it a decent amount at my last job to write test reports that had a lot of similar text with minor changes.
I also use it for dnd to help me quickly make the outlines of side characters & flesh out my world.

Porn has been ruined by AI too. Jokes aside it's really a boner killer.

Idk who faps to that whack shit but it's trying so hard to make everything look baby silk smooth with unrealistic bodies most likely stolen from hentai.

Internet search, e.g. Google, is now functionally almost completely useless. I use ChatGPT basically as a Google replacement.

I will still search for stuff - I use Kagi - but give up after half a dozen results if none of them are relevant and go to ChatGPT instead. Often, ChatGPT is more helpful. But sometimes it just makes a bunch of nonsense up.

ChatGPT is great for when you need to find something where you kind of know at least the vague shape of what you’re expecting and you have enough expertise to filter out any of the lies it makes up.

Im suprisingly on board for ai art. It does allow you to create whatewer you want without having the technical ability to do so. ( For example of you want a sick wallpaper ) Ot significantly lowers the floor as far as creating anything art related goes.

I'll use it to write scripts for repetitive tasks at my job. I never learned or know code so it's actually super helpful in that sense but that isn't really what OP is asking i don't think. I use AI by going on to their platform and initiating the interaction. I disable every form of AI I am capable of disabling/uninstalling. Every integrated sense of AI has been obnoxious.

People I talk to find it useful for front-end development

Yes, but they're the same sort of people that think Elon Musk is a genius.

For me throwing a graph in and telling it to create a table from it and stuff like that is really super helpful, since I often have to do this, and by hand it's a very tedious job. Sorting and cleaning tables and translating stuff is super handy and I use it quite often. But other then that I don't care.

The applications of what you call ai are absolutely limitless. But to be clear what you're calling "AI" isn't AI in terms of what you might want it to be what you're referring to are large language models or LLM's. Which aren't ai, not yet.

It's short sighted statements like this that really get my blood boiling.

If humanity actually achieves artificial intelligence it'll be the equivalent of the printing press or agriculture. It'll be like inventing the superconductor or micro transistors all over again. Our world will completely change for the better.

If your interactions with these llms have been negative, I can only assume that you have a strong bias against this type of technology and have simply not used it in a way that's applicable for you.

I personally use llms pretty much daily in my life and they have been nothing but an excellent tool.

Text generation is Frozen Yogurt now.

Noticeably worse, but you can have so much more.