[Vicinity9635]

If you're not using it where it's useful to you, then I still wouldn't say you're getting left behind, but you're making your job harder than it has to be. Anecdotally I've found it useful mostly for writing unit tests and sometimes debugging (can be as effective as a rubber duck).

It's like the 2025 version not not using an IDE.

It's a powerful tool. You still need to know when to and when not to use it.

[marcosdumay]

> It's like the 2025 version not not using an IDE.

That's right on the mark. It will save you a little bit of work on tasks that aren't the bottleneck on your productivity, and disrupt some random tasks that may or may not be important.

It's makes so little difference that plenty of people in 2025 don't use an IDE, and looking at their performance from the outside one just can't tell.

Except that LLMs have less potential to improve your tasks and more potential to be disruptive.

[Draiken]

You're right on the money. I've been amongst the most productive developers in every place I've worked at for the past 10 years while not using an IDE. AI is not even close to as revolutionary as it's being sold. Unfortunately, as always, the ones buying this crap are not the ones that actually do the work.

Even for writing tests, you have to proof-read every single line and triple check they didn't write a broken test. It's absolutely exhausting.

[nyarlathotep_]

I've encountered LLM generated comments that don't even reflect what the code is doing, or, worse, subtly describe the code inaccurately. The most insidious disenchanting code I've ever seen has been exactly of this sort, and it's getting produced by the boatload daily now.

[namaria]

I really don't understand what is going on. I try to, I read the papers, the threads, I think about it. But I can't figure this out.

How can it be that people expect that pumping more energy into closed systems could do anything else than raise entropy? Because that's what it is. You attach GPU farms to your code base and make them pump code into it? You're pumping energy into a closed system. The result cannot be other than greater entropy.

[marcosdumay]

Hum... In theory the closed system includes a database with most of the humanity's written works, and the people that know how the thing works expect it to push some information from the database into the code. (Even though, I will argue that the people that know how the thing works barely use it.)

The reason LLMs fail so often are not related to the fundamental of "garbage in, garbage out".

[cubano]

Of course! But just look at all the pretty lights all that energy is flashing.

[ryandrake]

Yea, "using an IDE" is a very good analogy. IDEs are not silver bullets, although they no doubt help some engineers. There are plenty of developers, on the other hand, who are amazingly productive without using IDEs.

[javier2]

I feel like most people that swear by their AI are also the ones using text editors instead of full IDEs with actually working refactoring, relevant auto complete or never write tests

[static_void]

Tests are one of the areas where it performs least well. I can ask an LLM to summarize the functionality of code and be happy with the answer, but the tests it writes are the most facile unit tests, just the null hypothesis tests and the like. "Here's a test that the constructor works." Cool.

[mrguyorama]

They are the exact same unit tests I never needed help to write, and the exact same unit tests that I can just blindly keep hitting tab to write with Intellij's NON-AI autocomplete.