Someone on lobste.rs recently asked what the “root cause” is for programmers feeling depressed about the introduction of LLMs to our industry. I of course can’t speak for all programmers, but something about the question motivated me to finally put my general feelings about LLMs into words. I’m reproducing my reply here, copied from the original.
This has been on my mind a lot. I want part of my job to be teaching people to be better software engineers. But I have serious reservations about LLMs, and it’s hard to imagine a company these days hiring someone to teach software engineering without expecting them to encourage heavy LLM use. I don’t have a good answer yet on how to resolve that tension. Maybe this post will get me a little closer.
As always, this post reflects my personal thoughts and opinions, not those of my employer.
For me, the root cause is that I see LLM use as unethical (mostly due to the environmental and intellectual property issues involved in training LLMs, but for other reasons as well). But much of the tech industry (including, it seems, the vast majority of employers) are saying “If you want to stay competitive and survive in this industry, you have to use LLMs to create software”. So the choice is seemingly (1) use the big unethical thing, or (2) leave the industry.
Coupled with that is the fact that I care deeply about software quality, but LLM-generated code is making many people care even less about quality (or notice it less) than they already did. There’s a general sense of “the LLM wrote it, so it’s probably good enough”. I deeply disagree with that sentiment, but making that argument is increasingly an uphill battle.
I do still enjoy the “programming” task itself, in the sense of writing code to do what I want, both at work and as an infrequent hobby. I’ve had to use agentic coding at work enough that I can see the value in it, and I definitely don’t feel a sense of loss or like I’m being replaced or anything.
If the ethical issues above were resolved, I’d be happy to use LLMs as part of (but not replacing all of) my work as a software engineer. But as it stands, I feel like most tech companies don’t share my values in the way that I thought we once did. So I feel stuck, and maybe even a little betrayed.
(Yes, there have always been ethical issues in tech, especially Big Tech, so maybe I was somewhat naive. But somehow this feels like a bigger, more overt shift in values than we’ve seen before.)