No. Experienced devs knew it would make tasks take longer, because we have common sense and technical knowledge.
I don’t blame randos for buying into the hype; what do they know? But by now we’re seeing that they have caught on to the scam.
Writing code with an AI as an experienced software developer is like writing code by instructing a junior developer.
… That keeps making the same mistakes over and over again because it never actually learns from what you try to teach it.
Yep, the junior is capable of learning.
Wait till I get hired as junior
Yeah, not all people who enter the industry should be doing so.
Most of this was boomers being boomers and claiming anyone and everyone should code.
This is not really true.
The way you teach an LLM, outside of training your own, is with rules files and MCP tools. Record your architectural constraints, favored dependencies, and style guide information in your rule files and the output you get is going to be vastly improved. Give the agent access to more information with MCP tools and it will make more informed decisions. Update them whenever you run into issues and the vast majority of your repeated problems will be resolved.
Well, that’s what they say, but then it doesn’t actually work, and even if it did it’s not any easier or cheaper than teaching humans to do it.
More to the point, that is exactly what the people in this study were doing.
If it’s doesn’t work for you, it’s because you’re a failure!
Still not convinced these LLM bros aren’t junior developers (at best) who someone gave a senior title to because everyone else left their shit hole company.
More to the point, that is exactly what the people in this study were doing.
They don’t really do into a lot of detail about what they were doing. But they have a table on limitations of the study that would indicate it is not.
We do not provide evidence that: There are not ways of using existing AI systems more effectively to achieve positive speedup in our exact setting. Cursor does not sample many tokens from LLMs, it may not use optimal prompting/scaffolding, and domain/repository-specific training/finetuning/few-shot learning could yield positive speedup.
Back to this:
even if it did it’s not any easier or cheaper than teaching humans to do it.
In my experience, the kinds of information that an AI needs to do its job effectively has a significant overlap with the info humans need when just starting on a project. The biggest problem for onboarding is typically poor or outdated internal documentation. Fix that for your humans and you have it for your LLMs at no extra cost. Use an LLM to convert your docs into rules files and to keep them up to date.
Your argument depends entirely on the assumption that you know more about using AI to support coding than the experienced devs that participated in this study. You want to support that claim with more than a “trust me, bro”?
Do you think that like nobody has access to AI or something? These guys are the ultimate authorities on AI usage? I won’t claim to be but I am a 15 YOE dev working with AI right now and I’ve found the quality is a lot better with better rules and context.
And, ultimately, I don’t really care if you believe me or not. I’m not here to sell you anything. Don’t use it the tools, doesn’t matter to me. Anybody else who does use them, give my advice a try an see if it helps you.
These guys all said the same thing before they participated in a study that proved that they were less efficient than their peers.
That is a moronic take. You would be better off learning to structure your approach to SW development than trying to learn how to use a glorified slop machine to plagiarize other people’s works.
Codex literally lies about being connected to configured MCP servers.
Are you trying to make a point that agents can’t use MCP based off of a picture of a tweet you saw or something?
This is why you use a downloaded llm and customize it, there’s ways to fix these issues.
Yeah, but LLMs still consistently don’t follow all rules they’re given, they randomly will not follow one or more with no indication they did so, so you can’t really fix these issues consistently, just most of the time.
Edit: to put this a little more clearly after a bit more thought: It’s not even necessarily a problem that it doesn’t always follow rules, it’s more so a problem that when it doesn’t follow the rules, there’s no indication it did so. If it had that, it would actually be fine!
Unless you are retraining the model locally at your 23 acre data center in your garage after every interaction, it’s still not learning anything. You are just dumping more data in to its temporary context.
Sounds like you have no clue what an LLM/AI actually is or is capable of.
https://medium.com/sciforce/step-by-step-guide-to-your-own-large-language-model-2b3fed6422d0
It’s not hard to keep a data library updated for context, and some are under a TB in siz.
Where are you getting your information from?
It seems you are still confusing context with training? Did you read that text and understand it?
Did you follow it yourself to build an llm?
Why do you think it’s solely a training issue?
So, you did not? Ok
What part of customize did you not understand?
And lots fit on personal computers dude, do you even know what different llms there are…?
One for programming doesn’t need all the fluff of books and art, so now it’s a manageable size. Llms are customizable to any degree, use your own data library for the context data even!
What part about how LLMs actually work do you not understand?
“Customizing” is just dumping more data in to it’s context. You can’t actually change the root behavior of an LLM without rebuilding it’s model.
“Customizing” is just dumping more data in to it’s context.
Yes, which would fix the incorrect coding issues. It’s not an llm issue, it’s too much data. Or remove the context causing that issue. These require a little legwork and knowledge to make useful. Like anything else.
You really don’t know how these work do you?
You do understand that the model weights and the context are not the same thing right? They operate completely differently and have different purposes.
Trying to change the model’s behavior using instructions in the context is going to fail. That’s like trying to change how a word processor works by typing in to the document. Sure, you can kind of get the formatting you want if you manhandle the data, but you haven’t changed how the application works.
But
All the fluff from books and art
Is not inside the context, that comes from training. So you know how an llm works?
If it’s constantly making an error, fix the context data dude. What about it an llm/ai makes you think this isn’t possible…? Lmfao, you just want to bitch about ai, not comprehend how they work.
This is Lemmy, bitching about AI is the norm.
Apparently some people would love to manage a fleet of virtual junior devs instead of coding themselves, I really don’t see the appeal.
I think the appeal is that they already tried to lean to code and failed.
Folks I know who are really excited about vibe coding are the ones who are tired of not having access to a programmer.
In some of their cases, vibe coding is a good enough answer. In other cases, it is not.
Their workplaces get to find out later which cases were which.
Funny cause my experience is completely the reverse. I’ve seen a ton of medium level developers just use copilot style auto complete without really digging into new workflows, and on the other end really experienced people spinning agents in parallel and getting a lot of shit done.
The “failed tech business people” are super hyped for ten minutes when cursor gives them a static html page for free, but they quickly grow very depressed when the actual work starts. Making sense of a code base is where the rubber meets the road, and agents won’t help if you have zero experience in a software factory.
That’s the funny thing. I definitely fall into the ‘medium level’ dev group (Coding is my job, but I haven’t written a single line of code in my spare time for years), and frankly - I really like Copilot. It’s like the standard code-completion on steroids. No need to spend excessive amounts of time describing the problem and review a massive blob of dubious code, just short-ish snippets of easily reviewed code based on current context.
Everyone seems to argue against AI as if vibe coding is the only option and you have to spend time describing every single task, but I’ve changed literally nothing in my normal workflow and get better and more relevant code completion results.
Obviously having to describe every task in detail taking edge cases into account is going to be a waste of time, but fortunately that’s not the only option.
Very true. I’ve been saying this for years. However, the flip side is you get the best results from AI by treating it as a junior developer as well. When you do, you can in fact have a fleet of virtual junior developer working for you as a senior.
However, and I tell this to the junior I work with: you are responsible for the code you put into production, regardless if you write it yourself or you used AI. You must review what it creates because you’re signing off on it.
That in turn means you may not save as much time as you think, because you have to review everything, and you have to make sure you understand everything.
But understanding will get progressively harder the more code is written by other people or AI. It’s best to try to stay current with the code base as it develops.
Unfortunately this cautious approach does not align with the profit motives of those trying to replace us with AI, so I remain cynical about the future.
Usually, having to wrangle a junior developer takes a senior more time than doing the junior’s job themselves. The problem grows the more juniors they’re responsible for, so having LLMs stimulate a fleet of junior developers will be a massive time sink and not faster than doing everything themselves. With real juniors, though, this can still be worthwhile, as eventually they’ll learn, and then require much less supervision and become a net positive. LLMs do not learn once they’re deployed, though, so the only way they get better is if a cleverer model is created that can stimulate a mid-level developer, and so far, the diminishing returns of progressively larger and larger models makes it seem pretty likely that something based on LLMs won’t be enough.
I’m a senior working with junior developers, guiding them through difficult tasks and delegating work to them. I also use AI for some of the work. Everything you say is correct.
However, that doesn’t stop a) some seniors from spinning up several copies of AI and test them like a group of juniors and b) management from seeing this as a way to cut personnel.
I think denying these facts as a senior is just shooting yourself in the foot. We need to find the most productive ways of using AI or become obsolete.
At the same time we need to ensure that juniors can develop into future seniors. AI is throwing a major wrench in the works of that, but management won’t care.
Basically, the smart thing to do is to identify where AI, seniors, and juniors all fit in. I think the bubble needs to pop before that truly happens, though. Right now there’s too much excitement to cut cost/salaries with the people holding the purse strings. Until AI companies start trying to actually make a profit, that won’t happen.
If LLMs aren’t going to reach a point where they outperform a junior developer who needs too much micromanaging to be a net gain to productivity, then AI’s not going to be a net gain to productivity, and the only productive way to use it is to fight its adoption, much like the only way to productively use keyboards that had a bunch of the letters missing would be to refuse to use them. It’s not worth worrying about obsolescence until such a time as there’s some evidence that they’re likely to be better, just like how it wasn’t worth worrying about obsolescence yet when neural nets were being worked on in the 80s.
You’re not wrong, but in my personal experience AI that I’ve used is already at the level of a decent intern, maybe fresh junior level. There’s no reason it can’t improve from there. In fact I get pretty good results by working incrementally to stay within its context window.
I was around for the dotcom bubble and I expect this to go similarly: at first there is a rush to put AI into everything. Then they start realizing they have to actually make money and the frivolous stuff drops by the wayside and the useful stuff remains.
But it doesn’t go away completely. After the dotcom bust, the Internet age was firmly upon us, just with less hype. I expect AI to follow a similar trend. So, we can hope for another AI winter or we can figure out where we fit in. I know which one I’m doing.
There’s a pretty good reason to think it’s not going to improve much. The size of models and amount of compute and training data required to create them is increasing much faster than their performance is increasing, and they’re already putting serious strain on the world’s ability to build and power computers, and the world’s ability to get human-written text into training sets (hence why so many sites are having to deploy things like Anubis to keep themselves functioning). The levers AI companies have access to are already pulled as far as they can go, and so the slowing of improvement can only increase, and the returns can only diminish faster.
I can only say I hope you’re right. I don’t like the way things are going, but I need to do what I can to adapt and survive so I choose to not put my hopes on AI failing anytime soon.
By the way, thank you for the thoughtful responses and discussion.
Someone on Mastodon was saying that whether you consider AI coding an advantage completely depends on whether you think of prompting the AI and verifying its output as “work.” If that’s work to you, the AI offers no benefit. If it’s not, then you may think you’ve freed up a bunch of time and energy.
The problem for me, then, is that I enjoy writing code. I do not enjoy telling other people what to do or reviewing their code. So AI is a valueless proposition to me because I like my job and am good at it.
People assumed X, but in one experiment the result was Y.
And in his many experiments the result was in fact X, if it was just 1 on which it was Y?
I don’t actually disagree with the article, I’m just pointing out the title is meaningless.
Not surprised.
In my last job, my boss used more and more AI. As a senior dev, I was very used to his coding patterns. I knew the code that he wrote and could generally follow what he made. The more he used AI? The less understandable, confusing and buggy the code became.
Eventually, the CEO of the company abused the “gains” of the AI “productivity” to push for more features with tighter deadlines. This meant the technical debt kept growing, and I got assigned to fixing the messes the AI was shitting all over the code base with.
In the end? We had several critical security vulnerabilities and a code base that even I couldn’t understand. It was dogshit. AI will only ever be used to “increase productivity” and profit while ignoring the chilling effects: lower quality code, buggy software and dogshit working conditions.
Enduring 3 months of this severely burnt me out, I had to quit. The rabid profit incentive needs to go to fucking hell. God I despise of tech bros.
Here’s the full paper for the study this article is about: Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (PDF).
The real slowdown comes after when you realize you don’t understand your own codebase because you relied too much on AI. To understand it well enough requires discipline, which in the current IT world is lacking anyway. Either you can rely entirely on AI or you need to monitor its every action, in which case you may be better off writing yourself. But this hybrid approach I don’t think will pan out particularly well.
And this gets worse over time because you still have to maintain it.
And as the cherry on top - https://www.techradar.com/pro/nearly-half-of-all-code-generated-by-ai-found-to-contain-security-flaws-even-big-llms-affected
No shit, Sherlock. Except that “AI” is a wrongly attributed marketing buzzword.
This study is over 6 months old, why is Fortune.com only writing about it now?
When writing code, I don’t let AI do the heavy lifting. Instead, I use it to push back the fog of war on tech I’m trying to master. At the same time, keep the dialogue to a space where I can verify what it’s giving me.
- Never ask leading questions. Every token you add to the conversation matters, so phrase your query in a way that forces the AI to connect the dots for you
- Don’t ask for deep reasoning and inference. It’s not built for this, and it will bullshit/hallucinate if you push it to do so.
- Ask for live hyperlinks so it’s easier to fact-check.
- Ask for code samples, algorithms, or snippets to do discrete tasks that you can easily follow.
- Ask for A/B comparisons between one stack you know by heart, and the other you’re exploring.
- It will screw this up, eventually. Report hallucinations back to the conversation.
About 20% of the time, it’ll suggest things that are entirely plausible and probably should exist, but don’t. Some platforms and APIs really do have barn-door-sized holes in them and it’s staggering how rapidly AI reports a false positive in these spaces. It’s almost as if the whole ML training stratagem assumes a kind of uniformity across the training set, on all axes, that leads to this flavor of hallucination. In any event, it’s been helpful to know this is where it’s most likely to trip up.
Edit: an example of one such API hole is when I asked ChatGPT for information about doing specific things in Datastar. This is kind of a curveball since there’s not a huge amount online about it. It first hallucinated an attribute namespace prefix of
data-star-which is incorrect (it usesdata-instead). It also dreamed up a JavaScript-callable API parked on a non-existentDatastar.object. Both of those concepts conform strongly to the broader world of browser-extending APIs, would be incredibly useful, and are things you might expect to be there in the first place.My problem with this, if I understand correctly, is I can usually do all of this faster without having to lead a LLM around by the nose and try to coerce it into being helpful.
That said, search engines do suck ass these days (thanks LLMs)
That’s been my biggest problem with the current state of affairs. It’s now easier to research newer tech through an LLM than it is to play search-result-wack-a-mole, on the off chance that what you need is on a forum that’s not Discord. At least an AI can mostly make sense of vendor docs and extrapolate a bit from there. That said, I don’t like it.
People will literally do anything to avoid rtfm
It’s a struggle even finding the manual these days if you don’t already know where it is / what it’s called. I was searching about an issue with my car recently and like 90% of the results are generic AI-generated “How to fix ______” with no actual information specific to the car I’m searching for.
I searched up a video to replace a part on my car. I did find it, but I also found 15 videos that were AI generated product reviews of the part.
I definitely also want my car parts to be “sleek and stylish” when hidden away under a plastic cover under the hood lmao
I like your strategy. I use a system prompt that forces it to ask a question if there are options or if it has to make assumptions. Controlling context is key. It will get lost if it has too much, so I start a new chat frequently. I also will do the same prompts on two models from different providers at the same time and cross reference the idiots to see if they are lying to me.
I use a system prompt that forces it to ask a question if there are options or if it has to make assumptions
I’m kind of amazed that even works. I’ll have to try that. Then again, I’ve asked ChatGPT to “respond to all prompts like a Magic 8-ball” and it knocked it out of the park.
so I start a new chat frequently.
I do this as well, and totally forgot to mention it. Yes, I keep the context small and fresh so that prior conversations (and hallucinations) can’t poison new dialogues.
I also will do the same prompts on two models from different providers at the same time and cross reference the idiots to see if they are lying to me.
Oooh… straight to my toolbox with that one. Cheers.
