Well, yeah, the problem with LLMs in a professional context is exactly that the second order effects of using them are so bad that they wipe out the direct effect gains (the time saving in creating the text).
It’s fine for entertainment if all you want is to be distracted for a while with some somewhat bland chewing-gum for the brain.
I’m finding it hard to reconcile the people who find it useless with those who find it useful. My own experience with languages is that it speeds me up for tasks I have the skills to verify, but with ones I don’t it can slow me down doing the verification and fixes.
I know software developers in a Fortune 500 company who swear that overall they are far more productive with “agents”. I don’t know if it’s a highly tailored system rather than the more all-purpose ones the general public have. So I do wonder if there are use cases for it.
It’s ruining the planet and our societies, though. And it’s distracting us from what really matters. Net negative for sure.
I have no idea if LLMs actually make me faster in the aggregate, but I’m more interested in using them to be better. For example, having a LLM review some code, or share a plan for and have it point out all of the flaws in your reasoning or alternatives you never considered.
I find it can speed me up, but it’s very easy to become lazy and find myself asking the AI to do things I could do perfectly well myself, quickly and using far less electricity. And then before I know it I have a huge pile of slop that I don’t know my way around, and I spend the time on the other end trying to understand and check it all. It feels amazing up front, but you pay later.
For corporate software, there’s so many bugfixes and features, which require so few additions/changes that it’s faster to do it than to write the prompt.
But obviously only if you’re able to hold the entire codebase in your skull, but a proficient developer should be able to do so.
If people begin outsourcing these changes to LLMs, then they’re cooked and no longer able to verify the output.
Yeah, it seems like a new type of judgement is needed that I don’t normally encounter in a day. Like, I use a hammer to hammer nails, then I put it down. I don’t try to cook with it. This tool has a weird, abstract nature to it where you can drift between using / misusing / abusing without necessarily noticing. A strange position to be in while trying to work at the same time .
Yeah strong black/white opinions on AI annoy the hell out of me. I find some AI tooling really helpful like fancy line completions and using conversationalLLMs as a last resort to solving problems, but I hate anything to do with agents or vibe coding in general if only just because I don’t want the field I enjoy so much being entirely replaced with being a project manager for a computer.
It makes me frustrated trying to talk about this stuff on Lemmy or other places though because it always ends up being two battling sides of “A SINGLE GENERATED CHARACTER MEANS THE ENTIRE PROJECT IS SLOP!!!” and “ANYONE WHO ISN’T VIBING OUT MILLIONS OF LOC WITH 100 DIFFERENT AGENTS GOING AT ONCE MAY AS WELL JUST KILL THEMSELVES!!!”
I can kinda get the sentiment behind that polarization. There’s such a strong drift towards LLM use that it takes hardline stances to try and slow it down
My code adjacent work is much faster, but programming itself is about the same.
So if I need to make a data visualization or UI for data input it’s way better, but if I have to do an analysis or build something it sucks all the knowledge out of the task and takes twice as long. At least. Because unless I’m forming a strong understanding of the information the task can’t progress.
it speeds me up for tasks I have the skills to verify
The issue is that the executives only see the potential to speed up, and not the necessity of having skilled people who can do that verifying.
Agents is just the LLM prompting itself with subtasks. This can be helpful to compartmentalize larger and more complex tasks. I find the tool calls most important. An agent that can compile and test its own code will make fewer mistakes (assuming the tests weren’t also written by LLMs but even then).
So there’s a self-correcting feedback loop that will improve quality, but is also a lot more expensive. The problem is that the temptation is huge to give more and more complex tasks to the agents and then they go off for 30min or longer and may or may not create complete and utter bullshit because something led them astray or they forgot something one agent did wrong and never fixed it. The end result may work, but it may be less efficient or come with a share of dead code that’s never called etc.
You end up with hundreds of lines to verify. You could just ask an agent to do it, but at some point adding more agents to the pile is not going to help… and probably takes longer than just reading it yourself.
It’s like a flamethrower, if you’re clearing brush it helps you go ten times faster. But if you dont know what you’re doing you’ll burn the whole property to the ground.
This is happening everywhere. My job is about 2 months behind schedule at the moment and theyre still pushing for AI solutions
Not a fan of the guy but I’m stealing “slop grenade.”
dw he stole much more
I’m not a fan of a lot of his politics but I honestly thought that interview was worth listening to. He has some interesting nuanced takes.
CEO Tobias Lütke… told employees last year that using AI is “a baseline expectation,” and they should first prove they “cannot get what they want done using AI” before asking for more resources.
Isn’t this an example of Play stupid games, win stupid prizes?
It looks like an example of Play stupid games, win stupid prizes.
Just playing uhm angels advocate? Not sure how to call it, the naive optimist perhaps!
Not necessarily: the statement (in isolation) is a call to learn and become capable to explain limitations. The criticized effect is basically “people not being able to explain the limits”.
I suspect that “finding limits” wasn’t incentived though (and I mean things like positive feedback, not money). And then you get shit-in-shit-out factories.
Or to rephrase your statement: If you want to play a game then stick to the rules you’ve set up - otherwise everyone’s allowed to cheat.
Ok, sure
However
In this case everyone that had a connection to reality told him that’s not a reasonable baseline.
You can force everyone to use it, but then you are working against that
Not exactly setting up your workforce for a win
I would love to find anything that had a more positive outlook … but alas perhaps I should ask OpenGrok because my creativity is not enough.
Fuck em.
A hammer makes a poor substitute for a knife & fork.
how else is ai slop going to be monetized? more slop more hallucinations more bubble.
The “purpose” for AI (the reason it is being FORCED on us) is that it allows “wealth” to access “skill” while at the same time keeping skilled people away from wealth.
NOW what billionaire would not want that?
What poor person would want that?
See the problem yet?
Defenestrate Mark
This currently only works for poor people because access to skill is free right now… and we see it not working for billionaires once the financial bubble around LLMs bursts.
Not that billionaires will suddenly be poor, but someone will be holding the bag eventually. We see efforts right now that the government/taxpayers will be the ones via bailouts. This is of course unacceptable but the patriotism argument that the West needs to outcompete China may work for some.
With CEOs - we need to switch to REAL grenades.
Billionaires do this to avoid income taxes.
The most infuriating part of this? Mr. Lütke endorses the idea of granting more votes to those who pay more income tax, including stripping the right to vote from people who do not pay income tax.
He has also called pensioners ‘dependents’ who shouldn’t get to vote:
New deal: when you get your pension deal it’s locked in and guaranteed. But now you are a dependent and that means no voting, just like dependents under age. Enjoy the deal, let people with a stake in the future decide.
https://moneywise.com/news/top-stories/shopify-ceo-tobi-lutke-pensioners-voting-rights
Well, considering I paid more in income tax than Bezos did, I think that’s a pretty damn good idea
You shouldn’t expect AI to generate a finished work product yet. If you’re forcing employees to use it, teach them to used it as a starting point and then put their own personal touch on it.
If you’re making a deck, have AI generate the bones, you add the details, then polish it up with AI if you want it to look fancy. If you expect it to turn a bunch of docs into a presentation only using prompts, you’re going to have a bad time.
Or, or - now hear me out - ban the use of AI to do a job that requires a human thought process.
Dude you’re getting paid to not have to use your brain. I finish my work in an hour and spend the rest of the day doing a bunch of corporate culture BS that gives you visibility to the bosses that actually decide your pay. Put effort where you get the most return. Hang out with your friends in the cafeteria and talk about WoW. Tell your boss you are in a mentor circle and slap that on your quarterly review.
Other than chatting with friends, that sounds like a nightmare. I’d rather be just doing the job that I enjoy and doing it well.
Now I think I’m the problem. Great.
Yeah bc that worked for the Amish Empire.
Weird comment considering the Amish are still around and, if anything, get proven right every day (except for the sexism and racism)
Get proven right about what? Plenty of Amish sects explicitly use technology if it’s in the course of their work or business. They use power tools for furniture making, have computers to handle payments, cell phones, trucks, etc.
They just don’t use it in their personal lives.
Correct, glad you agree.
Aw, now kith
The problem in my experience isn’t as much the work forcing it, it is the people who are very enthusiastic because they didn’t know what to do and now they think the AI does it for them, but they still have bad ideas.
Think of the person who has this “great app idea” but just needs a developer to write it. Well now these messes are having GenAI churn out… something. They don’t know if it works or not, but now all they need is someone to help finish off the “essentially done” work, or to accept the pull request the AI generated for your project.
Similar story for the creative content like videos and text. These people were limited by not having enough care to bother to learn or invest the time, but they thought they had amazing ideas. Now we see their uninspired slop bright to life in painful time killing form. The ones that might use it as just an accelerator but care about the output are drowned out by the “prompt and it’s done” folks that just don’t care.
Great, can we now abort this asinine “AI” bullshit and go back to normal GPU prices?
No, we’re going to continue on this path until we have to suck the liquids out of your body to cool the data center
They are hyping themselves. So is Fortune.
We all told you it was vaporware and you wanted to ride that bubble. Don’t pretend any of this was a surprise.
What I really want to know is, were the people putting out workslop, putting out anything better BEFORE AI. I mean, maybe they are just using a different tool to suck.
I think instead of stating that their performance was going to be measured by the quantitative measure of how much AI they used, instead of the qualitative measure of output, while making AI available to them, the CEO might have seen different results.
It’s almost like malicious compliance. “Hey everyone, do this.” Then they all do JUST that. What would you expect? LOL.
In my experience in a way it was better, because they had very low volume of output. Yes, pretty bad already but the GenaAI turned to their volume to 11.
This is one of the biggest downside, the worst people get the most volume out of it because they don’t know or carevto curate the output.
Looking at my last job before I left when AI was being pushed heavily I think I can see what you mean actually.
Before the incompetents would look at something and pass it on to second line without thinking. Now they copy paste the idiot AI outputs. It makes their stats look way better, wastes customer time much more and doesn’t help when it finally gets to me after 2 weeks of bullshit.
Customer satisfaction would be higher if you didn’t use it and did nothing. You are a net drain to the company now because you are trying to do things you don’t understand because of blind reliance on LLMs.
Heh, reminds me of an email thread. Someone stated a fairly straightforward, clear, and actionable course of action.
Then someone followed up with a GenAI email basically trying to be a sycophant to the person who wrote the plan and agreeing with them.
A VP did not get excited by the person that actually plainly declared what to do, but did announce that he was going to set up a meeting with the sycophant because he displayed so much insight, despite only repeating the first guy in a more confusing and buzzwordy way.
The AI generates corpo-speak in a very exciting way that further exacerbates an existing problem of promoting the wrong people.
“What separates workslop from low-quality work done by humans is that workslop looks legitimate on the surface while lacking the components that would make it useful.”
Lacking optional things like the truth/verifiable fact, substance/depth of any kind, a basic contextual understanding of the shared reality of the broader team or vendors, an actual point of view beyond the status quo and any creative thought or truly new ideas.
I prefer the term “Meat Proxy”, but I get that it’s more of an ad hominem. I use both terms.
You asked for it, you got it.
