MariahStrap in, kids, this is my hot take on AI. I want to start out by saying that I didn't use any sort of artificial help when I wrote this. This is all me and it's going to be long so grab a coffee. As most of my friends know, I have a love-hate relationship with AI, and said relationship influences my thoughts on it; a lot, so let's break it down. First, the hate. I'm still salty about being replaced by an AI audio editing service and losing a steady gig, even though I understand why I was phased out. From a financial standpoint, it made sense to replace me. Why pay me $100 per podcast episode when an AI can do my work for $15 a month. I know I can do better, but I also get that it's hard out here. I hate the fact I was so easily replaced, and I hate the fact that people are using AI irresponsibly and that everywhere I turn it's being shoved in my face like some sort of religion. I hate the fact AI is being asked for medical advice. It's googling your symptoms all over again just worse and good god why. I hate that people who might not be as in the know about technology as I am are being fooled by it. I hate that my facebook feed is crammed full of these AI stories that quite frankly make me want to vomit because there's only so much feel good recycled bullshit I can take, and people are feeding into that, too. I hate that my browser even uses AI and I have to sift through a lot to find a simple answer to a question. I despise ai-written articles. I hate the influx of AI-generated music that features in my spotify discovery weekly and release radar playlists. I said in a post a bit ago that yes, some of it is cool, but most of it is actual garbage. I hate that my PC might soon be AI-driven. No thank you. Get out of here with that mess. Now for the love, because there is some. I love that I can use AI to get an idea about pictures and videos of my almost 1-year-old son from his dads. I know that AI can be wrong, so I use resources, like his dads in this example to tell me if my descriptions are wrong, or if there's any detail that the AI might have missed. I am very aware that not everybody has resources like that, but I do, and I am grateful for them. I love that I can use AI to read the text of something like a card. I shared a story a while back about how I was with my brother at his friend's place and we played a game called mixtape, where you have cards with different scenarios on them that you have to pick songs to fit. Think apples to apples for music lovers like me. Rather than my brother or anyone else reading my cards for me, I was able to do it myself. I of course asked if I got the correct info because again, AI can be wrong, but for the most part in this case it wasn't. That was liberating and I had a great time playing that night. I love that I can use AI to help me get my thoughts together if I'm writing something out. With careful prompting, I can have a model ask me questions about the topic I want to write about and prompt me for my thoughts, then generate something after all my thoughts are down on the virtual paper. Once generated, I take what it gives me then fix it to match the way I write and correct any errors I find because I fact check everything. I *do not* let AI write something for me and call it mine. I'm fair at writing and I take pride in that and I'll be damned if I'll let an AI model take that away from me, but I do recognize when my thoughts are disjointed and when I just need prompts and questions to get me on track. It's a tool, friends, nothing more. I love that people are using AI as a tool to make more mainstream games accessible. I don't love that some folk don't check the AI's output, but if you do and you have the knowledge to fix issues with code then I think you're doing it right and I applaud you. It's not just games where accessibility with AI is a thing, either. Let's be real, the world is not made for folk like me or anyone with a disability, really, so if AI can assist in making something accessible? great. Again, using it as a tool is fantastic and I have no issue with that, and that extends to things like music and the process of composing and making a track. I think it's fine to take and have something like suno generate a track and then you as the musician build your own track around what you got given by the AI. Make it your own. I don't have a problem if you're just using suno for your own entertainment, go for it, be free, but if you're going to put something on a streaming service? Find some way to make it your own because let's be real, there is something clinical about AI-generated music even though some tracks are cool. I'm all for tools in your toolbox, but like anything else, if you're not being responsible about how you use said tools that's where things get dicy. AI is here to stay whether we like it or not, but it's up to us how we use it, and if you don't? That's fine too. There's a fine line and some folk might choose to step over it and some might not, and it's okay. Be aware of the risks, and remember that it's artificial intelligence. It's not a substitute, but it can be a tool. If you read through all of this I would love to know your thoughts. Have a good Monday, yall.
@MariahL I agree with every word of this. I will continue to say that I don't have a problem with AI. I have a problem with the "slop" and people refusing to use AI responsibly in public contexts. AI is a tool. Like any tool, no matter how primitive or refined, it can and will be misused. Such misuses are not the fault of the tool IMO, rather they are the fault of the people committing the misuses. I will never discount the fact that AI is changing some areas of life for the worse, taking jobs, promoting misinformation, being used for pulling in vulnerable people, and for that yes, I acknowledge these uses of AI are bad. But I will also never be able to fully hate AI because it's also done plenty of good things, and given me personally some independence where I would've previously had to ask sighties for help every 7 seconds. You wanna generate fake articles or songs or whatever for entertainment? Fine, go for it. Not my business what you do with AI in your personal time. You wanna talk to it, use it for music generation, make your own images or whatever, go ahead, whatever, I'm not going to come down on your head about how horrible you are just because you dared to touch AI. I just think people need to be more careful with it in the public sphere and stop overflowing everything with AI slop. I'm about to get massacred by someone I'm sure, but that's my take.
Talon@Kaliah @MariahL one of the biggest problems, but by far not the only one, with LLM's and modern so called "AI" is the people pushing it. Don't gotta look far into Sam Altman to realize that guy is a problem on two legs. Don't gotta look far into the people behind Suno and their investors to realize that those people are also extremely problematic. We already know about Facebook and Google and the list just kinda goes on like this. And use it or not, I think it is very important to keep this in mind. These people do not want to make the world a better place.
Robert Kingett@talon Also this "tool" will go away the next hype cycle comes around. Obviously, machine learning and the general tech will not go away. Machine learning will continue to advance but I do think after the bubble bursts we will see a lot more local and even specialized local LLMs. It is not here to stay, because investors won't allow it to stay. The above posts buys into the hype, unfortunately, otherwise I would have boosted it. I do use an LLM because I just can't get away from it, honestly, like in the Be My Eyes app and similar, but 2 truths can be in one's head at the same time. To the original poster, it is a complete accident an LLM just happens to make things more accessible, which is unfortunately why Enshittification will hit the most marginalized users the hardest when the tech starts to rot, which is already happening.
@WeirdWriter I do believe the general technology is here to stay since we do have open implementations of the fundamental algorithms. Whether that's a good or bad thing is a debate in and of itself, but the way these things work is out there. So while Anthropic or OpenAI might not necessarily stick around, language models/transformers I believe will. That said though even beyond those kinds of tools, I've bene thinking about accessibility and what it means. For example with the situation the US is putting itself in right now, the thought crossed my mind what I would do if the EU suddenly did not have access to GPS, if I couldn't use my iPhone, etc. So yeah I think the people who will suffer the most are the ones who most rely on access tech. As usual. Ugh. It annoys me thoroughly.
@talon Oh absolutely! Like I do believe after the hype cycle bursts we will see a lot more local and open source LLMs.