I mean I literally "tried it again" this morning, as a paying Copilot customer of 12 months, to the result I already described. And I do not want to "try it" - based on fluffy promises we've been hearing, it should "just work". Are you old enough to remember that phrase? It was a motto introduced by an engineering legend whose devices you're likely using every day. The reason why "everyone", including myself with 20+ years of experience is looking to do not "fun stuff" (please don't shove words into my mouth), but cool stuff (=hard problems) is that it produces an intrinsic sense of satisfaction, which in turn creates motivation to do more and eventually even produces wider gain for the society. Some of us went into engineering because of passion you know. We're not all former copy-writers retrained to be "frontend developers" for a higher salary, who are eager to push around CSS boxes. That's important work too, but I've definitely solved harder problems in my career. If it's boring for you and you think it's how it should be, then you are definitely doing it for the wrong reasons (I am assuming escaping a less profitable career).
Steve Jobs may be a legend at business, but an engineer he is not. To say nothing of the fact that whole reason "it just works" is because of said engineering. If you would like to be the innovator that finally solves that, then great! Otherwise you're just bloviating, and by god do we already have enough of that in this field.
I'm approaching 20 years of professional SWE experience myself. The boring shit is my bread and butter and its what pays the bills and then some. The business community trying to eliminate that should be seen as a very serious threat to all our futures.
AI is an extraordinary tool, if you can't make it work for you, you either suck at prompting, or are using the wrong tools, or are working in the wrong space. I've stated what I use, why not give those things a try?
The point is not the individual tools, which at this point are just wrappers around the major LLMs. The point is the snake oil salesmen of major LLM companies have been telling us for several years now that it is "just about to happen". A new technology revolution. A new post-scarcity world if you will. A tremendous increase in technological output, unleashed creativity etc. Altman routinely blabs about achieving AGI. Meanwhile hallucination of the models is a known feature, unfortunately it's not a bug we can fix. The hallucinations will never go away, because the LLM models are advanced text generators (quoting Charles Petzold) producing text based on essentially probability that one token should follow the next one. That means mate, you can be a superstar at the "advanced" skill of "prompting" (i.e. typing in conversational english language sentences), the crappy tool will still produce output that does not make sense, for example type out code with non-existing framework methods etc. Why? Because with every prompt, you retrain and re-tune the model a little bit. They don't even hold the authority of a dusty old encyclopedia. You use several tools simultaneously, why? Because you cannot really rely on any of them. So you try to get a mean minimum out of several. But a mean minimum of a sum of crap, will still end up crap. If any of the the 3-4 major LLM engines had any competitive advantages, they would have literally obliterated the competition by now. Why is that not happening? Where is the LLM equivalent of nascent Google obliterating Altavista and Excite or an equivalent of Windows 95 taking the PC, React taking over the web frontend etc? And by the way, you know that there was another famous guy at Apple, right?
They've been saying that kind of shit about everything AI related since fuzzy logic was the next big thing. It will never happen. AI will be used to cut staff and increase the workload of those remaining. The joke is on you for being susceptible to their hype.
I use a couple of different tools because they're each good at something that is useful to me. If Jetbrains AI service had a continue.dev/cline like interface and let me access all the models I want I might not deviate from that. But lucky for me work pays for everything.
You also seem awfully fixated on Copilot. How much exactly do you think your $12/month entitles you to?
Well thanks for confirming, you're getting "something" out of each, i.e. minimising mean error, because none of them is the ultimate tool. Copilot price is actually $19 per seat and running my own company, I pay a bit more than $19 bucks, you know for my employees, people like yourself. Why I am fixated on a single tool? Because each of those "tools" are wrappers around one of the major LLMs. I am surprised you don't know that. Copilot, Windsurf, CLine etc. are all just frontends for models by anthropic, google and chatgpt. So the output cannot be, by definition very different.
There is lots of value to be added in wrapping those tools. I am very well aware of what these things are. LLMs are not a fire-and-forget weapon, even though so many of you business types really really really want it to be. I mean jesus you sound almost as delusional as my bosses.
Business type? I am nothing near a business type, with two technical degrees and 20 years of hands-on experience. But I managed to build my own stable business over the years, in part due to being analytical and not rushing to conclusions, especially not over strangers on Internet ;) Where did you get the conclusion that I am delusional? It's actually the business types who think that these tools are magic, mind-blowing, etc. I am, like many other "technical types", pushing for the opposite view - yes to some extent useful, but no where near the magic they are being advertised as. Anyone who calls them "mindblowing", like some guys in my comment thread are either inexperienced/junior or removed from the complex parts of the work, perhaps focused on writing up React frontends or similar.
There is no hubris. The LLM technology has actually been existing for at least two decades, it's not some sudden breakthrough we suddenly discovered. And given the many billions of dollars it has sucked in, it's definitely a pile of crap. I have been a paying customer of Github Copilot for at least a year. Since the google search has been completely messed up, sometimes it can be useful to look up some cryptic error message. It can also sometimes help recall syntax of something. But it's not the magic machine they've been touting, it's definitely not the AGI, and my god, it's very prone to errors and to anyone who admires this tech, please for the love of god double- and tripple-check the crap that they generate, before you commit it to production. And by now it definitely feels like the 'self-driving' cars 'revolution'. They have been 'just around the corner', for like what, 15 years now?
No, LLMs have not exited for two decades. What a stupid comment. Millions of people are spending thousands of dollars because they're recieving tens of thousands of dollars of value from it.
Of course it's impossible to explain that to thickheaded dinosaurs on HN who think they're better than everyone and god's gift to IQ.
Please read more carefully.The LLM technology exists for at least two decades, well actually even more.You know the technology the LLMs are based on (neural networks, machine learning etc). I am not sure if after smartphones, the LLMs will now further impact intelligence of people like you. And take note: I have been one of the early adopters and I am actually paying for usage. My criticism comes from a realistic assesment of the actual value these tools provide, vs. what the marketing keeps promising (beyond trivial stuff like spinning up simple web apps). Oh by the way, I peeked into your comment history. I see you're one of those non-technical vibe-coder types. Well good luck with that mate and let us know how is your codebase doing in about a year (as someone else already warned you). And if you have any customers, make sure you arrange for a huge insurance coverage, you may need it.
Without bad intent, I am not sure I am even able to make sense of your sentence. Honestly it sounds as if you fed my comment into an AI tool and asked for a reply. Here is a tip for the former junior dev turned nascent GenAI-Vibecoding-manager - if you want to attack someone's credibility, especially that of an Internet stranger you are desperately trying to prove wrong, try to use something they said themselves, not something you are assuming about them. Just like I used what you said about yourself in one of your previous posts. Otherwise the same thing will keep happening over and over again and you'll keep guessing, revealing your own weak spots in domains of general knowledge and competence. My second advice to a junior dev would have been to read a book once in a while, but who needs books now that you have a magic machine as your source of truth, right?
I’m sorry but you will be the first to go in this new age. LLMs today are absolutely mind blowing, along with this image stuff. Either learn to adapt or remain a boomer.
Go where mate, to Sam Altman's retirement home for UBI recipients ? I studied neural networks while you were still a "concept of a plan" in the mind of your parents and unlike you, I know how they work. As one of the early and paying adopters of the technology, it would have been great if they worked as advertised. But they don't, and the only people "to go" will be idiots who think that a technology that 1) anyone can use 2) produces un-reliable outputs.. 3)..while sounding authoritative makes them an expert. Guess what, if both you and your buddy and your entire school can spin up a website with a few prompts, how much is your "skill" worth on the market? Ever heard of demand and offering? To me looks eerily similar to how the smartphones and social networks made everyone "technologists" ;)
> is that it produces an intrinsic sense of satisfaction, which in turn creates motivation to do more and eventually even produces wider gain for the society.
Which society? Because lately it looks like the tech leaders are on a rampage to destroy the society I live in.
Copilot doesn't use the full context length. Write scripts to dump relevant code into claude with it's 200K or the new Gemini with even more. It does much better with as much relevant stuff as you can get into context.
But I don't want to write additional scripts or do whatever additional work to make the 'wonder tool' work. I don't mind an occassional rewording of the prompt. But it is supposed to work more or less out of the box, at least this is how all of the LLMs are being advertised, all the time (even the lead article for this discussion).
LLMs are also primarily promoted through the web chat interface, not always magic wonder tools. With any project that will fit in claude/gemini's large context you use those interfaces and dump everything in with something like this:
(tree Source/; echo; for file in $(find Source/ -type f ) ; do echo ======== $file: ; cat $file; done ) > /mnt/c/Users/you/Desktop/claude_out.txt #claudesource
Then drag that into the chat.
You can also do stuff like pass in just the headers and a few relevant files
(tree Source/; echo; for file in $(find Source/ -type f -name '\*.h' ; echo Source/path/to/{file1,file2,file3}.cpp ) ; do echo ======== $file: ; cat $file; done ) > /mnt/c/Users/you/Desktop/claude_out.txt #claudeheaderselective
You can then just hit ctrl+r and type claude to refind it in shell history. Maybe that's too close to "writing scripts" for you but if you are searching a large codebase effectively without AI you are constantly writing stuff like that and now it reads it for you.
Put the command itself into claude too and tell claude itself to write a similar one for all the implementation files it finds it needs while looking those relevant files and headers.
If you want a wonder tool that will navigate and handle the context window and get in the right files into context for huge projects, try claude code or other agents, but they are still undergoing rapid improvements. Cursor has started adding some in too but as subscription calling into an expensive API they cost cut a lot on trying to minimize context.
They also let you now just point it at a github project and pull in what it needs, or tools build around the api model context protocol etc. to let it browse and pull it in.
No thank you for obviously good intent on your side, but I am not looking for scripting help here, nor am I a business type who does not code themselves. I just don't want do this when I am already paying for the tooling which should be able to do it themselves, as they already wrap Claude, ChatGPT and whatever other LLMs. And unless you're professionally developing with Microsoft stack, I'd advise to ditch the Windows+MinGW for Linux, or at the very least, a MacBook ;)
The tooling you are paying for doesn't work with the full abilities of the context so you need to do something else. Doesn't matter what it's supposed to do or that other people say it does everything for them well, it works a lot better with as much in context as possible on my experience. They do have other tools like RAG though in cursor, and it's much quicker iteration, ultimately a mix of what works best is what you should use, but not just block stuff out out of disappointment with one type of tool.
I am lucky in the sense that neither myself nor my business depend very much on these tools because we do work which is more complex than frontend web apps or whatever people use them for these days. We use them here and there, mainly because google search is such crap these days, but we had been doing very well without them too and could also turn them off. The only reason we still keep them around is that the cost is fairly low. However, I feel like we are missing the bigger picture here. My point is, all of these companies have been constantly hyping a near-AGI experience for the past 3 years at least. As a matter of principle, I refuse to do additional work for them to "make it work". They should have been working already without me thinking about how big their context window is or whatever. Do you ever have to think how your operating system works when you ask it to copy a file or how your phone works when you answer a call? I will leave it to some vibe-coder (what an absurd word) who actually does depend on those tools for their livelihood.
> As a matter of principle, I refuse to do additional work for them to "make it work". Do you ever have to think how your operating system works when you ask it to copy a file or how your phone works when you answer a call?
Doesn't matter, use the tool that makes it easy and get less context, or realize the limitations and don't fall for marketing of ease and get more context. You don't want to do additional work beyond what they sold you on, out of principle. But you are getting much less effective use by being irrationally ornery.
Ok now think about this in terms of items you own or likely own: What would you do if I sold you a car with 3 doors, after advertising it as having 5 doors instead? Would you accept it and try to work around that little inconvenience? Or would you return the product and demand your money back?
