don't bury your head in the sand
I think it's a bit hard to ignore the current state of the world when it comes to AI. Regardless of where go, there's a very strong pushback against AI (specifically, LLMs). And while we could go on and on about the pros and cons of AI, the drawbacks to education and impacts to society and all that, I wanted to rant a bit today about one thing that particularly gets me - and that's people burying their head in the sand.
I find it difficult to have constructive conversations about AI with people shoot off phrases like, "it's just the next crypto!" or "we'll hit model collapse and it'll all be useless!". Hyperbole offers nothing to the conversation; clearly, something has shifted society wide. While there are clear issues with the technology, it's simply illogical at this point to say that it's a fad, or that it will go away; to believe so is burying your head in the sand.
an anecdote on software engineering
Out of personal experience, I find this particularly severe in the junior software field; while I (somewhat luckily) entered the industry years ago, many of my juniors who are graduating now or soon are facing somewhat of a dilemma when it comes to AI - will it make their degrees worthless? Will the future of software engineering be one principle-distinguished-senior-staff engineer prompting 1000 instances of Claude? Some react in a very hostile way; of course it just produces slop! You'll see in a few years when all these companies shut down! 1 I would only work at place where all code is hand written!
Of course, some of these platitudes don't just come from juniors. But clearly, vibe-coding/agentic engineering/slop-coding has become somewhat mainstream. Anthropic couldn't have hit peak revenue without engineers talking to Claude all day long. And while we've seen some rises in downtimes and incidents, shouldn't we still take note when engineers say "I haven't really been coding in months"? At my job, even the more hard-headed engineers (die hard users of neovim) are still using Claude on the side, to explain unfamiliar code or speed up test writing.
Perhaps that's what it is - a tool. Tooling that can be used well or poorly, with the side effect of numbing your brain as you go. Either way, I think it's hard to deny that LLMs have somewhat embedded themselves into the day to day work of engineers; it remains to be seen if that pattern can be repeated in other industries (I fear it may be easier than we think).
the wrong message
I think my main problem with refusing to acknowledge (not admit defeat - simply acknowledge) LLMs is that the wrong lesson is being taken away. You should not ignore LLM; you should work despite it, against it, whatever you may choose. You should write by yourself, you should draw by yourself. 2 If you don't need augmentation when it comes to your daily life, you should continue working on yourself. But don't ignore its' existence. It will be there, and it will have an impact on your life, potentially negative; and you should be ready for it regardless.
a little afternote
- I don't want to come off as an AI-bro so please let me know if you think I did
- this was written as a bit of a single session rant so I hope it wasn't too rambly
as always I'm contactable by either guestbook (for things you'd like the public to see) or email at jottednotes@proton.me for feedback