Fuck AI
"We did it, Patrick! We made a technological breakthrough!"
A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.
Am I the only one who tried a LLM like twice, saw it gave out bullshit, then never tried it again (though I do see it forced on me with general searches, be it google or DDG or Bing or kinda anywhere now)?
I tried Copilot to answer a couple of coding questions, and I ended up having to take as much time to double check the answers/code it didn't seem worth it.
FYI, you can use noai.duckduckgo.com to disable AI searches :)
just wish they would put that (and the non-"please install our browser") URL into Firefox's dropdown of search engines!
Yep, I see very little value in them. I have techie friends who keep telling me I should try it more, but it just pisses me off and creeps me out. It took a long time getting this brain working as well as it does, and it's already headed back downhill. My neurons need the exercise.
I've asked LLMs a few simple coding questions, only to question why the code isn't working receiving an Oh sorry here try this malformed garbage instead.
I've asked Gemini to generate an image, only to receive a completely black image. When asked to describe the image Gemini would tell me what I was requesting it to generate. I saved the image, uploaded to Gemini and asked to describe the image. It gaslit me and straight lied that it was not the same image.
What a waste of time and resources.
People who know what they're doing don't use these "tools", it's the fools and morons out there who have no idea how anything works that use these things and they don't double check them before they implement their crap code.
I use it to help me write because of my dyslexia. But that’s really all it’s good for. Nothing advance like creating one line diagrams for electrical installs. I tried. It sucked at it.
Looks like learned helplessness is back from the grave with a new definition
It never went away.
Sincerely, someone who works in IT support.
Imagine asking AIs to form your own opinion.
"Grok, should I support Ukraine or Russia?"
Who would have thought that we are all becoming like Batman's villain, Two Face, to make decisions from an inanimate object.
The nickname is new, the behaviour isn't.
At the risk of coming off judgemental of family, my mother in law is exactly like this; especially when it comes to anything related to the computer.
My partner is an only child and every time mother dearest has any kind of issue, real or imagined, with her computer, she's hitting the speed dial for my partner's phone. At this point she's kind of in a mindset of "I've tried nothing and I'm all out of ideas". If she can't get an immediate response she actually starts to think about the problem logically, and tries to fix it herself.
Luckily, she hasn't been inundated with AI chat bots yet.
I'm certain that if she could manage to get to chat gpt, she would be asking it what to do about everything under the sun... Lucky for her, I work in IT support and manage what updates she gets, or more accurately, doesn't get.
She's a fairly mild example since she actually tries when she can't get an instant response from someone on what to do. There's plenty of people that do not.
I'm almost entirely convinced that some of the willful ignorance is simply people aggressively keeping to their job descriptions (at least when it comes to what I normally have to deal with)... They don't try to fix their computer because that's not their job. Even if they know how, they won't. That's not in their job description. It gives them an excuse to work less and get paid for it.
Regardless of the reason, people often know nothing about things and don't care to be informed on the subject, they just want the easy answer as to what they need to do next. Unfortunately for them, life is rarely that "black and white".
How soon till cult mechanicus is recognized for religious tax breaks?
Praise to the Machine God.
LLM thumping slopvangelicals.
those frickin clankers should just unplug themselves
We’re cooked.
I use Mistral Ai instead because I don’t touch and US products or services. The most common thing I ask is: check and tidy up these notes add formatting for readability, present as a markdown code block, add some nerd fonts.
I used to do this a lot with google, but now that it's barely functioning at giving me correct results... Yeah I use chatgpt a lot. I'd love it if there was a functioning search engine instead, but there isn't
Use duck.ai instead of chatgpt. Not as good. But it is private.
I will back up that recommendation with Lumo, a privacy focused AI from the makers of Proton Mail.
The only way I use chatgpt now is A: on the tor browser and, B: with a darknet email. however the free and more anonymous stuff (such as Duck.ai) is not THAT far behind. They use older models, but as times passes they will get better.
I've found to be duck.ai ok. The reasoning model is sometimes ok for basic code and math
I'm genuinely curious, what do you find chatbots useful for?
Honestly I literally use it as a google replacement. Idk what happened with google but I just can't get anything useful out of it at all, and so I'll just ask ChatGPT to search online for stuff if google proves useless.
The only other use I like for it is learning new languages, as a developer it can help a lot, as long as it doesn't make shit up, which it does fairly regularly tbh