Particularly, AI may starve the development of open source libraries. Which, ironically, will probably increase the need for employed programmers as companies accrue giant piles of shoddy in-house code that needs maintaining.

[–] VoterFrog@kbin.social 12 points 9 months ago (4 children)

Why do you think AI will starve open source?

[–] kibiz0r@midwest.social 17 points 9 months ago (3 children)

The amount of code I’ve seen copy-pasted from StackOverflow to do things like “group an array by key XYZ”, “dispatch requests in parallel with limit”, etc. when the dev should’ve known there were libs to help with these common tasks makes me think those devs will just use Copilot instead of SO, and do it way more often.

[–] VoterFrog@kbin.social 4 points 9 months ago (1 children)

I think that undersells most of the compelling open source libraries though. The one line or one function open source libraries could be starved, I guess. But entire frameworks are open source. We're not at the point yet where AI can develop software on that scale.

[–] kibiz0r@midwest.social 2 points 9 months ago

I agree wholeheartedly, and I think I failed to drive my point all the way home because I was typing on my phone.

I'm not worried that libs like left-pad will disappear. My comment that many devs will copy-paste stuff for "group by key" instead of bringing in e.g. lodash was meant to illustrate that devs often fail to find FOSS implementations even when the problem has an unambiguously correct solution with no transitive dependencies.

Frameworks are, of course, the higher-value part of FOSS. But they also require some buy-in, so it's hard to knock devs for not using them when they could've, because sometimes there are completely valid reasons for going without.

But here's the connection: Frameworks are made of many individual features, but they have some unifying abstractions that are shared across these features. If you treat every problem the way you treat "group by key", and just copy-paste the SO answer for "How do I cache the result of a GET?" over and over again, you may end up with a decent approximation of those individual features, but you'll lack any unifying abstraction.

Doing that manually, you'll quickly find it to be so painful that you can't help but find a framework to help you (assuming it's not too late to stop painting yourself into a corner). With AI helping you do this? You could probably get much, much farther in your hideous hoard of ad-hoc solutions without feeling the pain that makes you seek out a framework.

load more comments (1 replies)

load more comments (15 replies)