this post was submitted on 03 Nov 2026
582 points (98.3% liked)

Technology

89506 readers
3350 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots


founded 3 years ago
MODERATORS
top 50 comments
sorted by: hot top controversial new old
[–] isekaihero@ani.social 12 points 1 day ago (1 children)

When the lawnmower was invented in 1830 it created the unrealistic expectation that all well-heeled properties have carefully manicured lawns with flat cut grass. Suddenly even people who had no desire or interest in gardening were expected to own lawn tools and do yard work to adhere to the new expectation. That expectation spread across the world and now lawnmowers and yard tools are a multi-billion dollar industry that consumes everyone's time for no good reason. It's just part of human culture now. Every year all the supposedly intelligent, free-willed humanoids living on Earth dedicate their scant time, wealth, and resources to cut weeds near their domiciles only because it is expected of them. It accomplishes no positive benefit. It is a net waste of time and energy. It is a purely harmful thing to do. But if you ask anyone about it, they will initially be convinced it's a good thing because they don't actually think about it with logic. They just keep doing it because for the past 200 years we have all been doing it.

It's very possible AI will turn out to be a similar thing.

[–] tough_cookies@lemmy.cafe 6 points 1 day ago

I fucking hate Big Lawn. I have friends who let their yard grow wild and it's awesome. The pollinators love it. My HOA and spouse won't let me do it.

[–] Kissaki@feddit.org 14 points 1 day ago* (last edited 1 day ago) (1 children)

“…Employees worked at a faster pace, took on a broader scope of tasks, and extended work into more hours of the day, often without being asked to do so,” the researchers wrote.

How is "faster" measured? What is faster?

Pace doesn't say much about completion, total, quality, or repeated or thrown away work.

load more comments (1 replies)
[–] betanumerus@lemmy.ca 8 points 1 day ago

"but at least they're causing more fossil fuels to be burnt and that's why we all exist in the first place" /s

[–] corvi@piefed.social 135 points 2 days ago (5 children)

I can anecdotally attest to this. My company has built their product in such a way that it is unusable without AI. There are some things it legitimately expedites, but there are more things that used to take me 10 minutes that now take an hour and a half. And then I have to go validate the output anyway.

[–] FlashMobOfOne@lemmy.world 47 points 2 days ago* (last edited 2 days ago) (2 children)

Yup.

My work uses MS Teams and Sharepoint, and MS is dumbing down both products as much as is feasible in order to force people to use CoPilot more. It's moronic, and I hate discovering changes to the UI that appear to be bugs and then finding out it was deliberate. (Like buttons and scroll bars that are invisible until you hover a cursor over them. Stupid.)

But LLM's were only ever useful if the validation of their output didn't outweigh the supposed time saved, and that is rarely the case.

Stupid business practices. Can't wait for a popping of the bubble.

[–] MalReynolds@piefed.social 7 points 1 day ago

if the validation of their output didn’t outweigh the supposed time saved

This is the key. LLMs are (somewhat, there are style and maintenance issues)) useful where things are easy to prove correct, like mathematics (there are theorem validity provers) or programming (does it compile, pass tests).

Most of the rest of the world is not like that and tasks require judgement, so people need to validate that the LLM didn't 'hallucinate' (bullshit anthropomorphic word for 'get it wrong') again.

In subtle domains finding if an error exists can take longer than doing it yourself, with a worse failure rate, especially as skills atrophy. Same thing happened on production lines with the introduction of robots.

[–] Grilipper54@sh.itjust.works 3 points 1 day ago (1 children)

Teams implementation of copilot was done poorly. I'll use it sometimes by clicking copilot in the upper right corner. Its 50/50 after I use copilot in teams that a window will cover up the left side where all my chats are and it won't go away. I have to actually right click and quit teams to get the chats with people back up.

Clearly there was minimal testing by Microsoft before they rolled out copilot in teams.

[–] FlashMobOfOne@lemmy.world 2 points 1 day ago

Clearly there was minimal testing by Microsoft before they rolled out copilot in teams.

Oh yeah. They rolled out an update a month or two ago that turned notifications for project tasks I had that I had flagged for higher importance, inexplicably. Absolutely infuriating stuff happening with their vibe coders.

[–] Decq@lemmy.world 25 points 2 days ago (2 children)

I'm probably ignorant, as i left the conventional workforce before AI took off.. But how does that exactly make you work longer? If the AI isn't finished by the time the shift is over, then it's tomorrows problem, no?

[–] evasive_chimpanzee@lemmy.world 55 points 2 days ago (2 children)

Most people in jobs like this aren't paid hourly, they are salaried and expected to get the job done. Usually the problem isnt necessarily the runtime of an AI, but the inefficiency of a workflow that brought AI into its center.

It's like walking a child through how to cook a meal. Maybe it's technically less "work" since they are doing the chopping, measuring, stirring, etc, but it takes longer since you have to take time to be explicit about all your instructions, and correcting mistakes, etc. The difference is a child will eventually learn how to do it themselves, while in practice, and AI probably won't since it likely relies on inputs that can't just be piped in (cause if they could, you'd do the job with regular programming).

[–] Decq@lemmy.world 17 points 2 days ago (11 children)

Most people in jobs like this aren’t paid hourly, they are salaried and expected to get the job done. Usually the problem isnt necessarily the runtime of an AI, but the inefficiency of a workflow that brought AI into its center.

Sure, but if the tools provided/mandated by the employer are inadequate to do the job more efficiently. That should be the employers problem, not the employee's.

You can't as management both claim the need for AI out of efficiency, and blame it at the same time for you not getting your job done in time because of it. But then again, how often do you come across competent leadership these days. The lemon can always be squeezed a bit more I guess.

[–] ParlimentOfDoom@piefed.zip 39 points 2 days ago

Management is not known for being rational in many workplaces.

load more comments (10 replies)
load more comments (1 replies)
load more comments (1 replies)
[–] Carl@anarchist.nexus 12 points 2 days ago* (last edited 2 days ago) (1 children)

My company recently shifted to a new time clock system, which uses AI for basic search functions. Search is also intended to be the primary way users navigate the site. The salesperson’s marketing was “oh it’s so easy, you just describe what you want to do, and the AI takes you there!”

I need to approve my time card each week. The page is literally titled “Time Approval”. But searching for “Time Approval” usually doesn’t yield any results. Sometimes I’ll get lucky if I search for something like “I need to approve my time card for last week” but even that isn’t foolproof. Like it literally can’t find the page I need every week, even when I search for the exact title of the page. And who the fuck wants to type out an entire sentence just for a basic site directory search?

The way to get to the approval page through the UI (by clicking through, without using the AI-powered search function) is like four obscure menus deep. This is something that every employee is expected to do every week, and they buried it like they were planning on making a treasure map. They intentionally made the UI unintuitive, so users would feel compelled to use the AI search. But then the AI search doesn’t fucking work, so I end up having to navigate the unintuitive menus anyways. I ended up just bookmarking the approval page directly in my browser, so I can log in and then use the bookmark to jump straight to that page.

load more comments (1 replies)
load more comments (2 replies)
[–] tauonite@lemmy.world 54 points 2 days ago (2 children)

And all that time goes to babysitting a text machine... What a life they're living

[–] mrgoosmoos@lemmy.ca 27 points 2 days ago* (last edited 2 days ago) (11 children)

this is what my company owners don't seem to understand.

productive employees don't enjoy that. if that becomes our job, will either do it poorly or leave, and then you'll just have shitty people doing it, with shitty results.

they've backed off a little bit, but they're still trying to push AI tools to do parts of our job that we are still 100% responsible for the outcome of. fuck that.

actually now that I'm thinking about it, fuck, it fucking pisses me off. like I told them that an application with major value (which also would be one of the easiest things to implement) is just a chatbot that reads the customer spec and you can ask questions, so instead of sorting through hundreds of pages across multiple different files and file types, the chatbot can find anything relevant and then you can look closer at it. do we have that tool? no. are they trying to push other tools on us that are half complete, can't possibly do a good job, and only have marginal time savings? yes. like yeah sure let's fucking let AI do an engineering drawing of a part, I'm sure it'll make it a very nice easy to read drawing that the machine shop will love, and all the senior people who are already reviewing everybody else's work will love to review constant junior level quality work. that's probably their favorite part of the job, reviewing shit quality work. and they love it even more when they know that the feedback they give won't result in somebody getting better at it, they'll just see it again tomorrow.

load more comments (11 replies)
load more comments (1 replies)
[–] Snapz@lemmy.world 18 points 1 day ago

Cleaning up the mess for a lesser final product.

[–] victorz@lemmy.world 35 points 2 days ago

Meanwhile I'm having a great time doing the actual fun part — the programming.

[–] Tollana1234567@lemmy.today 16 points 1 day ago (1 children)

thats because 90 mins is spent re-reviewing the AI-generated tasks/codes.

[–] Kissaki@feddit.org 3 points 1 day ago (1 children)

My colleague: I created an AI skill for the review.

[–] mushroommunk@lemmy.today 1 points 1 day ago

Me: that explains why I've rejected it three times for failure to meet our coding standards

[–] RememberTheApollo_@lemmy.world 25 points 2 days ago

Probably because they fired 10% of rhe workforce based on promised AI usage returns on investment.

[–] PattyMcB@lemmy.world 68 points 2 days ago (1 children)

But they said we wouldn't have to work!

[–] Flower@sh.itjust.works 17 points 2 days ago

Well yes, that's why there are layoffs too.

[–] Armok_the_bunny@lemmy.world 16 points 2 days ago
[–] stopforgettingit@lemmy.world 23 points 2 days ago

Sounds about right. Since we started having to incorporate AI, my work load has increased but my production has stagnated, because I am having a hell of a time getting out of my own way now that I have to wade thru mountains of AI regurgitation. YAY progress!

[–] 7rokhym@lemmy.ca 22 points 2 days ago (1 children)

They fired 90% of my co-workers. So the people left are working longer hours to make up for the bad management decisions that assumed AI would magically make the things happen without people.

In reality, much less is getting done and the few people left are either working much harder with AI tools (or not), the rest are checked out looking for another job and don't appear to be active AI users.

[–] Kissaki@feddit.org 2 points 1 day ago

Are people still invested to that degree? I'm surprised they don't reject significantly or long term more work and let management feel consequences.

[–] rumba@lemmy.zip 20 points 2 days ago (2 children)

I'm in one of those rare jobs that AI helps immensely. That said, JESUS FUCKING CHRIST make it stop. I'm working WAY more than I used to and it's all fucking nailbiting levels of risk.

Just let the job market come back enough I can slide into something not so AI friendly

[–] Strider@lemmy.world 3 points 1 day ago (1 children)

Could you tell which field that job is in?

What I've seen so far is a lot of people basically getting good starts with something AI but mostly still having to correct too much so that either the prompt takes more work than the actual task or the work thereafter is more.

And I have seen some situations where llm also really made sense and was helping in general, but that was far and in between.

[–] rumba@lemmy.zip 2 points 1 day ago (1 children)

Anthropic enterprise models are very strong at following runbooks and managing infrastructure. To frame this properly, I spent about a month documenting the environment and creating runbooks. I can currently do things like the following:

Follow the steps below to roll over this cluster. {

Choose a server in this cluster by ID in numeric order (this server)

Use Ansible+vault to create a backup AMI of (this server)

Use Ansible+vault to collect health data on (this server) according to the server health runbook

Use Ansible+vault to disable notification for this server

Use Ansible+vault to preform a full kernel update and reboot the vm according to the update runbook.

Use Ansible+vault to collect health data on this server according to the server health runbook post update

Compare the pre and post health and server alarms for (this server), if there is an issue, pause, SOS notify oncall

Use Ansible+vault to enable notification for this server

}

Each runbook contains exact steps required, how to verify success, each one of them is able to pull the E-stop, freeze on that step and SOS notify.

[–] Strider@lemmy.world 2 points 1 day ago (1 children)

Ah, thank you for elaborating!

Though I have to say, this seems to be a thing I also created tooling for in the past.

So what I did for clusters was create an optional script where servers could drop a healthcheck if they're member of a cluster.

The idea was that they are able to check their cluster status, do actions and check if they're OK again.

So basically a simple framework for what you've described and that also was long before llm.

I'm not saying this to discredit, but to add more perspectives.

[–] rumba@lemmy.zip 2 points 1 day ago (1 children)

You're absolutely correct, 100% It's entirely scriptable. And I have done that work many, many times in the past 15 years

The interesting differences:

I didn't write the ansible jobs, it did to my specifications, i just had to review them. It can generate an Ansible job that would take me 30 minutes to properly write/document in less time than it takes me to review its output.

It knows how to do AWS AMI and RDS images without details. If Amazon changes something and it errors making the ami, it will, without intervention, find amazon docs and update it's process.

My playbooks are essentially Word documents (md tho) telling it to make an AMI backup and disable notifications using the monitor server who's name and IP are documented the neo4j. I don't tell it where the logs are, it knows. It'll fire up a Cloudtrail to hunt something down if it needs it.

The diffs on the health checks aren't binary. If it sees something it can't explain, it will go back further in time, inspect other nodes, even start going down the debugging route for the app, (as long as I give it docs on how it should debug). It doesn't miss a line. It doesn't blindly ignore things. I was doing a rebuild on a server the other day, and during the post-health check, it noted that a certain error on this node is present on 5 other nodes and that the most common cause is a minor config issue. I asked it to diagnose; it found a config file from the old image that we never touch. I've been running with that issue on 6 servers for years. By my hand alone, I'd never have gone back, researched it and fixed it. It's truly trivial, but it's also things like that I don't have the time to invest in that make this so useful.

And it goes through this level of detail for every node. It only bugs me if it finds something outside of scope that's unexplainable.

I don't give it any passwords or hard config. It queries Amazon or Neo4j, which are kept up to date programmatically.

The real terror is I can do this on several different projects at once. As my documentation, guardrails, and automation get better, I'm getting multiple times my own hours' work done every day, and I was really big on automation and self-healing stuff before this ever came along.

Could the node classifier go bonkers and wipe out everything? It could. But I'm careful not to just cron up a bunch of jobs and take a nap. Probably the next step would be to make some overseers to verify backups are being made and it's waiting between jobs.

[–] Strider@lemmy.world 2 points 19 hours ago

Thank you for providing more context and substance for your use cases since it's something I have not experienced.

Keeping an eye on changes especially is always a pain.

[–] Grilipper54@sh.itjust.works 9 points 2 days ago (1 children)

LLM help at my job too but only if the employee has the knowledge to correct the issues.

Today I needed to send an email with a few details, the most important being the customers email needed to be included in the body of the email.

I tell copilot that. What does it provide.... Everything except the customers email. I was able to prompt it again to include that information but 90% of my team would've just sent the original copilot response instead of rereading it.

[–] Kissaki@feddit.org 2 points 1 day ago

LLM help at my job too but only if the employee has the knowledge to correct the issues.

My problem colleague is the major beneficiary of our AI. I'm sure it makes them even worse, quality and quantity.

Worst part is we employed them as a supposedly already experienced dev. He's way worse than all juniors I onboarded, guided, and who improved.

[–] wonderingwanderer@sopuli.xyz 15 points 2 days ago (1 children)

And some troll was arguing with me a month or so ago about how technology supposedly reduces workloads.

I kept explaining to them how any increases in efficiency only lead their employers to give them more work to do. The laborers are still milked for the value of their labor, in some cases for even longer hours. The only difference is the employer benefits from the increased production.

They seriously tried arguing that the industrial revolution was somehow good for textile workers. Like no, traditional craftspeople got run out of business because they couldn't keep up with the sweatshops that employed children in dangerous conditions for absurdly long hours for very little pay with no days off, and all just to create a soulless product that could be sold in larger quantities.

AI is no different, except that it doesn't even increase profits for the employers.

load more comments (1 replies)
[–] NonconfrontationaI@lemmy.ml 3 points 1 day ago
load more comments
view more: next ›