Ask Lemmy

I liked AI before GPT3

It was fun playing with a terrible image/word generator

I played around with it when it was a lot more abstract. I remember wombo had some neat art styles you could create images in. I didn't realize home much shit was plageriszed though

I've been a luddite since long before AI. AI is black box engineering. It's a shield they can and do use to create a malicious product.

As you pointed out, the most obvious use case is reducing cost of labor regardless of whether total labor is reduced.

I didn't like the idea of generative AI ever, even less when it was marketed as a replacement for creativity. But I DID use it once, tryna see if I could do that thing some people did to manipulate it into getting me free game codes. It never worked

I liked it when all it did was help you remember what options to use with tar to unzip a tarball.

Nope never liked it. Never will.

It doesn't matter if you like it or not.

We should think about what to do with the current AI hype (or bubble), or How a better AI should be made, etc.

I really liked Deep Dream. As a fan of Lovecraftian horror, it really tickled a niche for me. When chatGPT first hit, I was down with the general vibe, but then gestures around at everything.

Never bought into it, would never say I liked it, I'm not adverse to it, or at least I wasn't until the USA government decided it had to be integrated into all policy making decisions.

Also I don't feel we've reached the level of "intelligence" , there's a lot of tricks that it can do, and it has vast databases to pull from, but it's not hitting me as intelligent, and the more I read of it's discoveries" the more convinced I am of my idea. The people using it are intelligent, and are coming up with neat ideas using it as a tool, but the models itself aren't coming up with anything without human coaxing.

I've heard it's much better to use ai as an editor regarding things you already know, than it is to trust ai to fill in the gaps of your knowledge.

Initially I was into it. I had a long ongoing conversation with cgpt helping me flesh out some ideas for a tabletop RPG. I was more or less trying to get randomized names and backstories. Although it did not loop, it eventually became repetitive. Also, every backstory came with a side dose of sunshine and glitter that wasn't always helpful.

I found it most interesting that it seemed to know the rules of Pokemon Tabletop, but when pressed with specifics it always invented information. It was almost always worthless at generating a character because it made up skills and abilities that didn't exist. The moment you attempted to apply logic and fit the pieces together, it all fell apart.

When it came to image generation, I created a few scenes and attempted to mash them together in gimp. By the time I was done, it was just as efficient as grabbing an image from the web and using that as my template.

The one thing I continue to love AI for is it is absolutely cracked at image recognition. Whats this plant? An aerial yam, duh! What's this shower head? Well, the original manufacturer is gone but the exact same part is still being sold without a label - here's the webpage we found a match on. That still amazes me.

But in general I hate how AI is just slowly rotting not just the internet and social media but just human talent and general. In 20 years we are not gonna have nearly as many young artists and writers because they have to compete with AI able to crank out slop for free. AI cannot (and in my opinion, almost by definition will never be able to) equal the very best artists and writers, but how are up and comers supposed to get to that level? Think of all the all time great creative people who started off just grinding as a cartoonist, background character, or pulp fiction writer. Those are gonna be the ones to go first. Instead of a, for example, 2 year pipeline to be able to scrape by pursuing a creative profession it could be a 10 year pipeline.

I used to loathe all things AI. I used to abhor AI coding agents, too.

I actually pushed back against some AI initiatives at some of my jobs.

Fast forward to now, AI does about 60%-80% of my work on a given day. It saves me boatloads of time. Coupled with figma and dev tools MCP servers, it sucks slightly less at CSS. I've cobbled together some pretty solid rulesets and tactics for context building that have recently let me tackle an entire month of work in ~2 days, meeting company best practices and code coverage and even gathering demo screen shots for me without me having to do it myself.

I probably spend about $500/month on various AI services now, but I'm cranking out at least 3x what I used to (and being paid commensurate to that output, only way to justify it). It can't last forever but I have succumbed, reluctantly at first, if only to get what I can from it before it blows up.

As for AI image gen, no. Video? No. Anything ChatGPT or Grok related? No.