As a software engineer, please do.
I’ve actively started doing tickets manually again. Just reviewing is exhausting, and yes, in part because I’m overwhelmed by what I don’t understand. I’m somewhat new in our codebase and letting AI do all the work means I’ll never get to know it in depth. That means I’m not meeting the productivity goals managers have already adopted, though. :/
Or forever.
My sibling has learned incompetence from AI. They show off the things they had AI make and how they never could have done that themselves when they are things they know how to do themselves (make simple graphics and presentations), and would take much less time than they say they spent refining the prompts.
Please turn it off… I am currently dealing with a
bunch of fuwell meaning but intellectually challenged individuals at an IT team who just reimaged a machine running the master database for one of their sites based on what a chatbot told them.Now what concerns me most is that this site operates critical life saving equipment. While our software isn’t too important, some minimum wage workers are about to have a very high stress weekend is about the worst that will happen and then early next week it should be back to normal. But the same IT team has access to the life saving equipment as well.
The whiplash between apparent anti-fascism and purity politics on this site is wild.
Right? Purity politics. I’m not sure what to make of it. I believe it’s mostly driven by profit. I haven’t read anything about anyone actually driving this as a general strategy. It makes little sense except for general disruption of social cohesion and it feels too spontaneous.
I believe it started with climate / science denialism and fascism, things that you cannot tolerate because they lead to intolerance. Things where the emotional abuse eventually got too much and people just categorically rejected anyone from the other camp. But this has been massively expanded now. Veganism, cars, vaccinations, conspiracy theories, diplomacy, AI.
Right so, this is definitely related to what “purity politics” means, but not what I was referencing. I’m talking about the (liberal) tendency on this site for people to supposedly challenge purity-based political systems and ideologies like fascism, white supremacy, and settler-colonialism one minute, and then accept “piece of technology corrupts the human soul,” as an explanation for something better explained through systemic analysis or cultural critique.
The idea that “people’s brains don’t work because of AI,” is deepy puritanical in nature, and attributes the harmful quality to the technology itself and not the socioeconomic conditions where what that AI produces is made subject to liberal capitalist imperatives and logics. Conveniently, it then takes those values for granted in favour of a narrative that is also markedly ableist in its message. Oh, do people who are disabled and require AI use to access ableist education institutions or information on the increasingly oblique and inaccessible internet “wreck their brains?” I wonder if a fascist would even disagree with that narrative.
The global economy is probably already dependent on some Claude code instance updating an Excel spreadsheet on someone’s laptop with a note ”don’t close the lid!” attached to it
When I was a teenager I used to work at an airport and they had critical infrastructure running on things like Basic and Assembly, to transfer data between the computers and modern systems you have to use floppy disks because the computers that they ran on didn’t have drivers to support USB and we couldn’t upgrade the computers because the software wouldn’t work on any other hardware.
So the AI laptop will be at home amongst all the ancient but supercritical junk that seems to be necessary to run our world.
At a previous job, I had to build laptops that would multi-boot DOS, Win95, Win98, and XP, all on one laptop. Our field-tech’s software had to have those specific versions of windows for the machines that they serviced, because of how the software interacted with USB and embeded floppy drives.
If you want to scare yourself:
spoiler
They serviced Medical devices: X-rays, MRIs, CT-Scanners, etc.
Honestly, I’d prefer those medical device run on older software than even Win10 lol
That’s dumb. The “work” in maintaining this excel file was the prompting and overall agent logic. In reality you’d be using something like openrouter and the instant claude “dies” for whatever reason you swap to whatever the freshest AI is. This never becomes ancient, it keeps up with the times.
I don’t know about that the entirety of European airspace was closed early in the year because of a pilot inputed some malformed data and it crashed both the primary and backup system. The backup system just fed the primary systems inputs into it, so that was inevitable.
I’m not sure what your point is. Garbage in garbage out, that’s just the law of computers. That would br the pilot’s mistake.
The point is there’s zero error checking.
That’s on you, when using AI you always have a verifier script afterwards, it’s basic common sense. And if a script can’t do it then human, if you’re saving 90% of the time instead of 100% it’s still fine.
How’s it my job to fix the aviation industry?
This is a perfect example of why I’ve never used AI and never will.
when using AI you always have a verifier script afterwards
LOL. Sounds like you’re wasting 110% time here.
I wonder if Proton can run these old softwares
Proton would be over the top unless they’re using graphics APIs or other common gaming related libraries.
That said, Wine/Bottles would be a possible solution. But so is a VM that runs the software in it’s native environment with pass through ports or even port device/emulation.
Linux /POSIX is great, but its not always the right solution.
Reminds me of some stupid company maybe a few months ago that had their PRs failing due to copilot being down the dumbasses didn’t build an “if no AI” failsafe into their workflow. And no one knew how to recover it
I know my work already has reports built by running data through Claude and having it spit out an xls. Though, it’s not quite as integral as the all-important closet laptop.
It’s a poison pill. We can say we want to turn it off to check to see if anyone can still think, but they’re inventing it to turn us all into mindless dolts, so, turning it off kind of defeats their purposes.
As a millennial the only positive of LLMs is that younger generations won’t threaten my career
Younger generations are, by and large, as technologically illiterate as boomers are. It doesnt apply to all of them, but the majority of them grew up with nothing but easy UIs. They may not chicken-pick to type, but they still cant troubleshoot anything and tend to know nothing outside of interacting with a typical interface. They definitely grew up with nothing like Windows 95 or 98 (or even XP) where you had to actually understand how a computer works to some extent to use the thing (or fix it)
Millennials, particularly the younger millennials especially, grew up right in the sweet spot to be at a tech advantage. At least as far as the average person goes. Obviously anyone passionate about tech will be far ahead of the average person, regardless of age
Wholeheartedly agree. This is from 2013, and I feel it’s only gotten worse since then.
Oh, I remember reading this … 13 years ago … . 👴
Arguably even windows 95/98 was rediculously easy compared to dos or Unix systems before it.
Sit a kid down in front of a c64 with a manual and see if they can do anything on it. If they can read the manual that is…
Sure these systems are “obsolete” but the brain training is priceless. Kids now are training their brains to love ads and corporations. Its gotten super bad. And I’m not even an old person (just been into tech history my whole life) so its rediculous I’m even saying this.
We tried to teach ours and got “dad can you just fix it I’m in a game”. Then no desire to learn in free time.
Would you mind sharing how you tried to teach and what you think went wrong?
When we built his PC we invited him to learn about it as we picked components and as we put them together. Every time he had a computer problem, we’d try to explain why it happened and how we fixed it. Just no interest at all unfortunately.
I think all the time about how a lot of boomer guys know how to work on cars real well.
This is because most of them took auto shop classes in high school.
As a millenial, I’m great with computers… because I learned how to use them in school. It appears that after I left school, educators were like “welp, kids know how to use computers now!” and stopped teaching kids how to use computers in school. This is both bonkers and mind numbing. I also wish I knew more about working on cars. Anyway.
The older gen z too I would say
As a millenial looking to change careers, because I was done with the old one. No, it sucks!
I agree. I am more than happy to pass the torch when the time comes. I certainly don’t want to work forever, but I also want to be able to train competent replacements at some point.
They will if you don’t use AI though
Don’t believe you, seems like useless nonsense to me
Considering how much competitive jobs have become refusing to use a powerful tool like AI won’t help. I’m not saying that you should use it just that people using it will be advantaged over these who don’t
I work in a creative industry, as far as we are concerned it is plagiarism
It is also non deterministic crap that developers should know better than to use
I work in a creative industry, as far as we are concerned it is plagiarism
AI can be useful in pretty much every field, if you are shoveling snow llms can help you making a plan and draw a route to do it faster
Why don’t you actually try this and see how it goes?
I work for a landscaping company, my boss tried to get us to use AI to develop site plans… we went back to doing it the old fashioned way because AI is fuckin useless for stuff like this.
give it a bunch of years and you will see taking it over any other industry, same as computers and mobile phones
We have path finding algorithms. Large Language Models aren’t suited for that.
Probably speaking to the choir here, but I do feel it is worth pointing out that those same algorithms were considered AI until like 3-5 years ago when the zeitgeist shifted to AI=Transformers/LLMs
I know enough to understand it’s useless, it will fall apart when the psychosis wears off
Video generators are becoming excellent and image generators already are, what makes you think people will give up on these tools? The industry is already adapting, 1/3 of the ads you see now are made with AI
We should turn off AI.
Can you also turn off LLM generated websites for that week? If so, deal.
You are assuming they worked to begin with.
AI has been an amazing tool for those with no ability to think, to operate slightly like a person who can.
Like a peg leg helps a amutee walk. The issue is they are using to run and claiming they are the fastest.
They weren’t using their brains before ai, what made you think they would now?
Are YOU using your brain right now?
The CEO’s brains are definitely broken. The rest of us? Mostly fine.
Something a friend and I were discussing today was that you can’t really avoid AI and remain productive right now. (We both teach university Computer Science.)
Ignoring how so much of modern society is running on it, you’d have to complete isolate to avoid it and it’s reach entirely.
Additionally, the tooling and infrastructure to support how things were done just a few years ago has all been stripped away. Online search is ass on every major engine right now, most of the forums are dead, documentation and discussion lives inside of private Discord channels. It feels like there just isn’t a back to go back to anymore.
People are severely overestimating the importance of AI. It’s more FOMO than anything else.
Wait until you have to pay non-subsidized prices for the use of LLM’s, how productive are they in reality when the real costs come in to play? You could just train and hire actual people to do the job far more accurately with some quality control?
Not denying there’s some fields with great benefits from LLM’s, but writing emails, answering phonecalls aren’t it. LLM’s cant even transcribe conversations with any proper accuracy and as a society we now trust them to make decisions.
It boggles the mind, really.
On the contrary, it does a great job with responding to work Emails! I paste in the chain and prompt it “How can I tell this person to fuck off in corporate speak?” And I haven’t been fired yet! ;)
‘Yet’ is doing a lot there.
People notice when they get an AI generated response. It ‘looks’ professional, but that’s not what it ever comes across as. It comes across as lazy and incompetent.
I mean, if you use the AI incompetently and lazily, then sure? I think if you use anything incompetently and lazily you’ll get that result. The trick is to not be an absolute dumbass and then you’ll be fine. Know what can and cannot do well, then use it where appropriate to take advantage of the 80/20 rule.
Hi, I work in the trades. AI is not a regular part of my life. I have never used any of these chatbots because they seem… half-cooked. Like an Early Access game on Steam. Anyway, the whole thing has zero relevance to my job.
“you’d have to complete isolate to avoid it and it’s reach entirely.”
Nah. I mean, obviously for you, teaching computer science, I can see why you feel this way. But lots of people truly never engage with this stuff because it’s pretty underwhelming. My main interaction with AI thus far has been telling my doctor not to use the chatbot that puts lies into people’s medical records, and that’s about it.
“Online search is ass”
Isn’t this, like, explicitly because of AI?
Here’s the funny thing … they never worked to begin with!
I cannot stress enough how correct this is. It feels like it’d be similar to early Covid lockdown. People with real jobs that contribute to things are still working while everyone else with the cushy meaningless jobs are sent home to do nothing where they complain and make things harder for the people actually doing things.
I literally quit my job to join them because I couldn’t take it anymore and was met with an insane amount of harassment. Apparently I had be lucky to be considered essential.
The Covid lockdowns made me feel crazy. There I was, sitting at home playing video games and smoking weed all day, every day, like God intended. Meanwhile, my roommate was going stir crazy because he really wanted to go back to work. I still don’t understand it.
That’s a good comparison, because it illustrates very well why LLMs are so (seemingly) successful. If this useful technology were used primarily by the people who actually worked during COVID, there wouldn’t be a problem, because these people are very good at assessing what’s possible and what isn’t - since they actually understand their work. The problem - and the absurd hype - arises simply because LLMs reinforce the completely absurd ideas of incompetent people - and these are the ones who had nothing to do during the pandemic, since they do nothing but schedule pointless meetings and waste the time of those who are actually doing something meaningful.
