1. Be civil
No trolling, bigotry or other insulting / annoying behaviour
2. No politics
This is non-politics community. For political memes please go to !politicalmemes@lemmy.world
3. No recent reposts
Check for reposts when posting a meme, you can only repost after 1 month
4. No bots
No bots without the express approval of the mods or the admins
5. No Spam/Ads/AI Slop
No advertisements or spam. This is an instance rule and the only way to live. We also consider AI slop to be spam in this community and is subject to removal.
A collection of some classic Lemmy memes for your enjoyment
cross-posted from: https://lemmy.blahaj.zone/post/46123456
That’s how I felt waking up this morning (as of posting). Who else can relate?
You're directing your hate the wrong way, the problem is how generative AI is employed by the greedy, not the technology itself.
How do you employ ai non-greedily, pray tell?
Depends on what you want to accomplish. How do you employ a hammer non-greedly? You can use AI for meteorological forecasts, predicting diseases, data compression, the uses are endless. For the case of generative AI, it helps writing code, generate synthetic data, create quick proofs of concept etc.
Greed is when CEOs force it's use to justify laying off 30% of the staff.
Non-greed is everywhere and self hosting open source models exists. You can use it as any other technology / tool for yourself (personal assistant, coding assistant) Think of how much faster medical research can be done with it. Language barrier use cases as well. 2 individuals that don't share a common language can have a conversation thanks to NLP.
It's just a technology that calculates patterns.
Fucking easily? "Claude, convert these Postman JSON collections to JMeter load tests with switchable environments based on this shared config file we already use for our existing JMeter tests" when you barely know Postman OR JMeter and you can still do it in half the time of a developer who knows both.
If that's another language to you, how about "Making people print this job application PDF, fill it out, scan it and email it to us is insane, old-timey silliness from the turn of the last century. Convert the PDF to a fillable form, and make sure the fillable fields are properly located between the lines of the field names they correspond to" and then it just does it and now applicants can apply online and you didn't have to pay an expensive human to do menial one-off work in applications they don't really know.
There's a ton of dumb, time-wasting shit you don't have to do anymore because you can just make the robot do it. It's not all just straight plagiarism and stealing artwork and dead Internet propaganda-bot garbage. You can use it for good or evil. The power generation and inefficiency are still unsolved problems, the super aggressive build-out arms race is pure idiocy and we absolutely need to slow this whole thing down, and I'll never forgive these greedy fucks for ruining the PC components market every few years, first crypto now this, but there are use cases that allow people to do things they couldn't otherwise do.
Properly used, AI is a force multiplier. Improperly used, it's a resource-burning race to the bottom, but if you can't think of legitimate uses for a general-purpose software robot you can tell to do tedious stuff in plain language and it'll just do it with a bit of guidance, then I don't know what to tell you.
convert these Postman JSON collections to JMeter load tests
Just use jmeter! Jmeter and corpo-curl are two different tools, it's really not that hard to design your own test plans.
I also don't understand making the receptionist do it, e2e tests are actually important and should be designed by someone familiar with the system they are testing.
Convert the PDF to a fillable form
You mean ocr? The thing that we already had an ai boom/bust cycle about in the 60s?! The thing the USPS has used to read chickenscratch addresses on all envelopes since the 70s?
Here, this project has been filling your exact use case for more than a dozen years now: https://github.com/ocrmypdf
and make sure the fillable fields are properly located between the lines of the field names they correspond to" and then it just does it and now applicants can apply online and you didn't have to pay an expensive human to do menial one-off work in applications they don't really know.
There's a ton of dumb, time-wasting shit you don't have to do anymore because you can just make the robot do it. It's not all just straight plagiarism and stealing artwork and dead Internet propaganda-bot garbage. You can use it for good or evil. The power generation and inefficiency are still unsolved problems, the super aggressive build-out arms race is pure idiocy and we absolutely need to slow this whole thing down, and I'll never forgive these greedy fucks for ruining the PC components market every few years, first crypto now this, but there are use cases that allow people to do things they couldn't otherwise do.
Properly used, AI is a force multiplier. Improperly used, it's a resource-burning race to the bottom, but if you can't think of legitimate uses for a general-purpose software robot you can tell to do tedious stuff in plain language and it'll just do it with a bit of guidance, then I don't know what to tell you.
Big wall of nothing, didn't read.
Well yeah. But also the fact that AI is run by (and therefore tuned to the greed and desires of) big tech CEOs is also part of the problem. I agree with you, but it’s a more complex problem than simply saying ‘people in charge are to blame for this mess’
Wrong. Slop generating and cognitive offloading is bad and even the most mundane use of local AI contributes to ecological collapse for nothing in return.
I don't see how offloading a task that is trivial and that I won't have much to learn out of it is "bad", in that case you shouldn't use calculators. Of course, that is different than completely letting an LLM think for you.
As for the ecological collapse, that is inherently a problem with capitalism and big tech, AI models don't decide the hardware or what power source they can be trained on.
I optimize industrial robots using AI tools to analyze data, which allowed us to work far faster and improve machines we otherwise couldn't. Each optimized robot usually saves hundreds of watts 24/7 for 10-20 years to come. The optimized programs also reduce wear, sometimes extending predicted service life by multiple years. But yeah, I'm sure that contributes to ecological collapse
You sure like the word optimized!
Wow, you're right: your single use-case fixed all the problems with global AI use.
Great work and my bad!
Also, absolutely zero people are criticizing maching learning when referencing AI, but you would know that if your comment wasnt slop.
What else did Grok tell you to say?
I never said it fixes all problems, this was in response to a post saying "all AI existing is bad". Tons of Anti-AI people don't even know the difference between LLMs and machine learning, and who says LLMs aren't a part of this? "Write a python script to plot data in this log file"?
But since you apparently know my job better than I do, feel free to apply, we're hiring!
Lose the wilful ignorance, dude
The fucking irony. I don't respect you and you will get nothing in return for your subservience, just so you know.
Subservience to what? Having basic common sense and not being drowned by hate is not subservience.
What am I being subservient to? Don't say corporations, I'd far rather run ai locally using open sourced and open weight models, or on duck.ai
You haters are dull, and less creative and more often wrong than the AI you so loudly hate
Corporations and the billionaires
WRONG. Try again? Better yet, don't
RIGHT. lol
Be the change you want to see. You are the leader so...
That’s why I aim to lead an anti-AI community
I wish this was the sort of problem I woke up with on my mind. Like I would kill to have your problems.
I also with this shit was the only thing on my mind, I really do. In this community, it is
Your performative hate is dull
Your trend-following buzzwords are worthless.