I’ve never wanted to punch a word in the face until the word “agentic” started popping up everywhere.
I refuse to even pronounce it correctly.
“ajanteek” is probably what I’d say if I was forced to utter such unspeakable trash
What we want is software that behaves predictably. Since LLMs don’t do that, we don’t want them or their “agents.”
For fuck’s sake.
To be clear, “normal” people are using LLMs like crazy. This I believe is specifically talking about is installed local agents. Boomers and non tech industry suits are using LLMs left and right and won’t shut up about them.
They are not using agents. Most people use LLMs as Google on steroids.
Google on LSD (the hallucinate results)
I actually really like the technology that has been collectively lumped together as “AI”. I think it’s fairly useful now, and suspect we’re at the beginning of the curve for this technology, and it will only continue to improve, even as it becomes more efficient. I understand this is a wildly unpopular stance in these parts, judging by the avalanche of downvotes I get for having this opinion.
What I don’t like is feeding tons of information about myself to a giant tech company. I run a local LLM on my phone and another on my (fairly decent) computer, and sometimes I use duckduckgo’s front end, which anonymizes prompts.
As for agents, it’s pretty much as you say. I just am not ready to have an LLM take action without my direct supervision, because I’m confident that between flaws in the LLM and flaws in how I give it direction, something will go sideways.
So you’re decided not to use a useful tool, because you don’t like it. Gotcha.
I once thought the fediverse was full of cool, techsavy people. Turned out it’s full of old people yelling at clouds.
I’ve had Gemini downright lie or completely fuck up before and it’s not a rare occurrence. That means you have to verify everything anyway and hell at that point maybe you should have just looked it up yourself.
First of: When? Few months make a huge difference. Development happens fast.
Second: Gemini is not the top of the line. You could do worse - if you use the fucking LLMs from meta.
I know Gemini isn’t the top. It’s all I’m willing to tolerate.
That doesn’t make any sense.:
„This sub-par product of an evil mega corporation that uses their market power to shove this product down everyone’s throat is all I’m willing to tolerate.“
As bad as Antrophic and openAI are, their product is better and they don’t embed it into phones and search engines. (Not saying they wouldnt‘t if they have the chance, but they haven’t)
It makes sense. I’m not spending a dime… I’m not signing up and giving my credentials out to any other AI … I already have Google . Gemini is good enough for the things I look up monus occasionally being dead wrong which all of them can do. I don’t make stupid memes or videos.
Yeah okay. I understand the monetary angle.
However I also have never spent a dime on AI, and I don’t use Gemini. I run Qwen3.6 on my graphics card, or use opencodes free tier. Deepseek-v4- flash is pretty good.
I’ve read a comment recently that Googles local Gemma models hallucinate less than Gemini, but I don’t know if that’s true, since I hardly use Gemma and Gemini not at all.
I recommend downloading some local models and a harness like pi or opencode if you want to play around an have the hardware for it.
Sir this is lemmy. Most people here just follow whatever the echo chamber tells them to. Thinking or referencing reality is typically not advised.
Because you don’t need AI when you’ve got NI.
I like Northern Ireland, despite its problems, but I’ve never considered it an alternative to AI. Interesting insight there.
Its pretty good at giving correct answers but god its slow and expensive. You gotta print and post like 2 million letters asking your question, wait for the concensus to be achived, wait for the comittee to conbein and compile the answer, then wait for the answer to come back.
They could at least streamline it a bit, maybe call instead of mailing the answer, and save a bit on the post by just having a dedicated radio signal that blasts the question to the entire population instead of having to send everyone a letter.
Don’t these fart-huffing jackasses understand that I actually need to read my e-mails to plan physical meetings with other flesh-based entities?
It’s difficult for me to fathom how someone could risk their work getting fucked up by non-deterministic software.
To me, it’s not worth the risk or the hassle to set up and babysit a piece of software that might get shit wrong when I can just do it correctly in the first place.
I’m the other fence. My daily reporting has 12-15 documents output, that I then previously had to spend 3-5 hours moving this item over to this form, and that item over to that form, and management wants everything to be kept locked PDF so no excel macros can be used… upload everything to claude, let it run. Sure… sure, itll make errors here and there… but no one actually checks what these documents pop out… only that they’re there. And that alone, has cut my workload by half. So, thanks AI for giving me my time back, and thanks boomer bosses, for not knowing how to do basic computer tasks badly enough they will never figure it out!
Using AI to complete bullshit work tasks that shouldn’t exist in the first place. Yep, I know that feeling very well.
Had loads in my last job. I could spend a few hours and struggle to fill in a few sentences for this bullshit form HR keep sending us, or I could ask ChatGPT to write the most obviously generic corpospeak crap that for some reason people like and are even impressed when I say that I used AI for it. Am I the only one that feels like it shows that the task is pointless?
Fundamentally, they cost way, way too much to run. LLMs are useful tools (I use them a bit for data analysis and coding help) and the people who use them for that kind of thing might pay the full cost in some form (not as individuals but their companies might). But they’re trying to push them into everything and people who can’t think for themselves and want them to help them write emails and such are not going to foot the bill. Not even close.
I think they’ve realised that forcing universal adoption is the only way that this could work, to get everyone hooked on it then jack up the prices. But regular consumers will never, ever be the ones paying the enormous costs of LLMs. Even companies will start to realise that there is no point paying huge amounts just for everyone to be able to outsource writing emails and basic tasks to LLMs. And short of any huge breakthrough, they are incapable of doing most jobs.
It’s not difficult to understand. This whole AI bubble is just a push to get enough people, a critical mass, over a threshold that they start paying for the usage. The Bubble was waiting for the money to run out which is what built the bubble. The money is running out and the subscriptions are ready for purchase.
The software development industry is ready to purchase subscriptions, but no other sectors it seems.
Maybe but even there there are limits on how much they will be willing to spend for junior level programming no matter how much of a ship fast fix broken methodology they adopt.
What junior programmers?
I mean this is quite biased article. My dad is the most basic user ever. Lately he’s thriving just using the simple google ai results and ai features in word. It’s fucking surreal to me a 76 year old man suggesting me to use ai and he says that it made his life way easier.
Ai is a tool. Hating or loving a tool is just fanboyism. Its no different than worshipping (or hating) photoshop.
I think there were already some research findings, how software developers tend to drastically overestimate how much AI helped them, while AI was actually hindering their productivity. It could be the same case here.
What AI is really good at is spitting out plausible sounding garbage, pretending it did some comprehensive research on the topic, and making people feel like it’s helpful without neccessarily being so.
Yeah its a great tool if your work is based on regressive data. What people think it can progressively develop new things and invent. Which I highly doubt it’ll be anything like that any time.
I see ai like Xbox Kinect. When it was first released it was like future of everything. Now we use it for Face ID. We just need the bubble to pop so companies can actually develop for the good of humanity
I know quite a few people who have bought into agents 100%. Its wild to hear them talk about letting an agent run for days on their personal computer, grinding away on some task. The way they talk about it, you would think they were babysitting a child or something.
What do they use them for? I have played with it a little out of curiosity, even got an LLM running on my phone until it got stuck in an endless loop spewing nonsense. Not touched agents yet.
I haven’t really found much use for LLMs yet beyond rewording or summarising text and even that is rarely worth the effort of loading it up.
Its mostly used for building large software projects or doing online research into some topic. They will use the LLM to create an outline of the work to do, the let agents figure out how to to all those steps, and how to check if the answer is correct or not.
Looked up some suggested uses as first projects and they sound so boring and useless. Smart home stuff, no thanks. Manage personal finances and budgets, don’t need tools for that let alone AI powered ones. Email, I don’t want to use email in the first place.
This article’s title is ridiculous. What do you mean ‘why don’t normal people use agentic AI?’ Were they supposed to?
We need 12 billion people to subscribe to ultra-premium-plus-plus-super AI accounts to deliver on the expected subscription revenue that is the basis for the stock price. The fact that many ordinary people aren’t paying AI agents to watch movies for them, and read the news for them and engage in deep research across a range of topics… They’re practically engaging in a boycott.
I wonder if it’s more like VR.
People don’t like losing control over their technology in reality.
People want to retain control over their technology, like it’s a tool.
Virtual reality removes awareness and agency in the real world.
Same as agents do.
What does this have to do with VR? VR doesn’t remove anything, it’s just a fancy monitor.
a fancy monitor that makes like 1/3 of the people who view it throw up.
Removes [your first step to be able to remedy a problem] your awareness of what’s going on around you in the real world. [Many people feel very unnerved by it because of this aspect.]
… I was going to share a link here to whatever first video I found of any of the many videos of thieves sneaking into people’s homes while they’re oblivious on VR, but someone has created a VR game called thief sneak, where the wearer of VR can play a thief sneaking into people’s homes, making it too hard to find the videos of that really happening. The potential for irony now, seems obvious…
Anyhoo, beyond that flooding-the-zone, plenty else goes wrong, born of that aspect, and more. https://www.youtube.com/results?search_query=VR+fails
I have agents force fed to me at work. I’ve already got my own honed workflows I’ve refined over the years; now, Outlook and Teams and Various agentic frameworks are spamming me with summaries of stuff I either already know or don’t need to know… except they still get it wrong constantly, so I can’t trust what they think I should prioritize (usually it’s work I’ve already assigned to others).
Claude Code agents are decent, but they need to create discrete work objects that can be properly reviewed, so act like junior coders.
The main issue I have with most agents today is that they try to replace MY work, making it more complicated and time consuming, instead of completing the work I can’t bother to do myself. And in order to do those things effectively, I have to loosen my privacy and security models to the point where they effectively don’t exist. Which I’d never do for another human, so why should I do it for an agent that’s more likely to catastrophically screw things up at record speed?
Good job putting to words one of the core issues I’ve had with using LLM chat for work. Stop trying to design an entire fucking system or solution and just answer the damn question I actually asked, by using natural language to search documentation and existing implementations.
I would tell it to only answer exactly what I asked, but sometimes I do get value from it “exploring” on occasion. It’s just so much damn worse with it than my newest junior co-worker who is known on our team for getting hours deep into tangential rabbit holes before coming up for air and asking seemingly nonsequiteur questions.
Back when search engines were becoming a thing, people would search more or less using natural language, e.g., “I want to see cat pictures and also cute kittens.” but early search engines performed poorly with that kind of query; it was much better to just type keywords, e.g. “cat kitten pictures”. Exacerbating the situation was the fact that search engines weren’t very good at the time, even if you used carefully considered keywords.
It might seem silly now, but using a search engine effectively was a skill people had to learn, and a skill not everyone had.
Using chatbots is the same thing, I think. They’re not super reliable to begin with, and on top of that, I don’t think people understand the importance of a well crafted prompt. The combination can make using them very frustrating.
I read somewhere that teaching people about how LLMs work is way better at protecting them from the harms of LLMs than trying to convince them not to use them.
This would include prompt engineering i guess! Which to be fair i do not know about.
The video makes fun of it by having the vibe coder put "… or you go to jail " at the end of every prompt. https://www.youtube.com/watch?v=JeNS1ZNHQs8
