A coder considers the waning days of the craft

Some people I feel fear losing their siloed prestige built on arcane software knowledge. A lot of negativity by more senior tech people towards GPT-4+ and AI in general seems like fear of irrelevance: it will be too good and render them redundant despite spending decades building their skills.

This is the conceit of knowledge work in general. You're only as relevant as much as you keep up.

It's why it continues to tend young.

The older you get, the more opinionated you get and with opinions come inflexibility so the less you are open to new ways of thinking, even those which may be superior.

Essentially people have criticisms of newer tools that have become industry standards as unnecessary and inferior.

Regardless of the merits of that argument, the new tooling is where the industry is at. The best hackers I know are extreme versions of this. Increasing dissagreability appears to be the crucial flaw of competency unless you work hard to avoid it.

The most extreme version of this I remember was about 20 years ago when I met a man of about 55 who was clearly a very brilliant computer engineer with impressive credentials that I have forgotten (a prestigious PhD and work history) but said he had always dismissed microcomputers, yes, as in not minicomputers and mainframes as toys and a waste of time. This was like 2002. I was still in college then and I vowed to never become that guy.

Still trying...

The stuff he describes should rightly be replaced by an AI. However, I think we can still level up. After initial excitement about GPT-generated code I realize that the danger is that we will very quickly generate a lot of code with subtle bugs that will take even longer to debug. If I write the code myself I (or whoever I have look over my shoulder) will be quicker finding the bugs. If it's GPT code I'll first have to dig through the idiosyncrasies, then find the logic error.

Also: writing multi-threaded or otherwise concurrent code is hard to imagine even with GPT 6 or 7.

I honestly wouldn't expect it to be the first application, when the technology eventually does get developed, and we have the power budget necessary to actually operate one of these units.

Until then, it's not a question of "eclipsed," it's a labor market, so it's a question of "efficiency" and the large language model people have a very long way to go to produce something that's generally as capable as a programmer.

So.. I guess if your craft is "scripting" you might be able to get by with GPT-4, but to imagine that your child is not going to ever need to "program" is eager nonsense.

i kind of agree but also it kind of sucks spending hours debugging code in which gpt-4 has carefully concealed numerous bugs

i mean raise your hand if debugging code that looks obviously correct is the part of programming you enjoy most?

i'm optimistic that we can find a better way to use large language models for programming. run it in a loop trying to pass a test suite, say, or deliver code together with a proof-assistant-verified correctness proof