mikemarsh

Yeah, like anything it's easy to go far in one direction or the other: for some organizations the opposite approach may be needed.

Maybe "filtering" is a better term than "shielding", since the latter assumes an adversarial relationship between "devs" and "suits" at the outset, and the terms we use even to think to ourselves about these things influence how we ultimately behave.

This is brilliant, excellent example of the ideal of "management as shielding devs from higher-level corporate shenanigans" in action.

Yeah, it's tough not being able to think for myself these days.

Hopefully this is a wake-up call for them to drop the AI hype-chasing and refocus on making the best IDEs for us dinosaurs who, for whatever crazy reason, still code by hand and appreciate solid (deterministic) tools to help us out.

On a very related note, I'm probably very late to this discussion, but has anyone else noticed autocomplete in JetBrains products has gotten significantly worse after the advent of LLMs? It would previously take into account type signatures and all kinds of context, now it's clearly just a very minimal next token generator, often producing code that doesn't even compile.

Replying to your reply: I actually wish I had watched way fewer movies (and fewer times) as a kid. I envy you!

"Some of you may die, but it's a sacrifice I am willing to make"

Digital tools are made even more not value neutral by the fact that other humans are actively influencing them even while you're using them, via "the algorithm", "dark design patterns", etc.

At least with an AK-47 or a hammer, once you own it physically, it's yours and the original creator isn't sending you subtle psychological nudges to utilize it one way or the other.

This article describes the behavior consumers of AI, but I think the far more interesting spiritual question is about the producers:

How many of those involved in AI genuinely believe some degree of the "If Anyone Builds it, Everyone Dies"-style narrative where "AGI" is a kind of omnipotent destroyer god that can also somehow be harnessed for "good" in the hands of "the right people"?

> No First-Person Output, No Apologies

This is precisely the appeal of AI though, and a key factor in influencing people's attitudes towards it. Why would any AI company want to stop this? (I realize the author knows this already)

> In the post-AGI economy, humans will eventually stop writing and reading code

What's the definition of "AGI" these days? I've heard everything from "sci-fi simulated consciousness", to "does really good on benchmarks" to "whatever makes OpenAI X amount of money".

Perhaps the definition in this specific case is circular, "whenever humans stop writing and reading code"?

Interesting, thanks for the recommendation!

Thus far I still am, but I'm in a very weird and atypical work situation at the moment.

Weirdness aside, I'm continuing to write code as I used to, despite the overall company _cultural_ push towards AI, given the lack of direct orders, and pray my output is good enough that the "powers that be" leave me alone.

I am also in the job search process, and am also still noticing plenty of listings that don't mention AI.

I remember seeing some JS cryptography library a long time (maybe 10 years?) ago that provided an async hashing function for exactly this reason: to split up what would normally be a simple synchronous computation and yield more often to the main thread.

Can't recall the name, but I remember thinking that was very clever once I understood why it was doing that, and am glad to see the concept explained directly here.

A key point that's always missed in this kind of discussion is that people in the past didn't worship comfort quite to the extent people do now (although of course they'd often prefer it).

Sure, a materialist could go back and psychologize any glimmer of satisfaction anyone in the past might have felt as some kind of "cope", but we should at least give our ancestors the respect of properly representing their hierarchy of values.