Hacker News Reader

Comment by lackoftactics | Hacker News Reader

This seems to extend even beyond the arts. I would say that the race to add more features to web and mobile apps will face a similar obstacle.

How long can we continue making 10x more features just because we can, until somebody realizes that users do not have the attention to spend on them? There might be a shift, but currently, we are too focused on pleasing shareholders.

View on HN

I think as a developer this is one of the biggest takeaways I've had from this whole thing. I used to view features sitting endlessly in a JIRA backlog as a bad sign; a signal that we weren't prioritizing the right thing. But in reality this serves as a natural selection mechanic; if devs don't care enough about it to get it done and PM doesn't see the value in it then it slowly fades away.

With agents you can just throw them at every JIRA ticket and get them (badly) implemented. There's no rhyme or reason if it's needed or even wanted, and spending time to think about it is 'wasted'. So the numbers on the burn down chart go up, but actual productivity is stagnant or becomes negative as unnecessary bloat sneaks into your app or game.

I'm a big fan of forgetting. I throw out the phrase "forgetting is the point" a lot, loosely paraphrased from some musician (I forget who!) that said if you couldn't remember the lyrics they must not have been very good. Somewhere in the original spirit of Agile there was no Jira board, because the next thing to do was obvious to the team, and if it wasn't obvious it was easily discoverable, just go ask some users what they need. The idea of writing down every idea and letting AI crap out the features is entirely antithetical to creating good software.

In his memoir Stephen King says not to write your ideas down since the good ones stick around anyway, and that a notebook is a way to immortalize bad ideas.

Paul McCartney maybe?

"We were writing songs that were memorable, not because we wanted them to be memorable but because we had to remember them"

Had to refresh my memory, but it was "If it ain't worth remembering, it ain't worth writing down.", from Ronnie Van Zant.

Damn, enshittification was already one of the toughest unsolved problems that successful software platforms face. It had not yet occurred to me that AI was going to increase the pace.

It's pretty clear to me that the primary value-add of AI is as an enshittification accelerator. (At least in the commercial sector.)

The primary value of AI is solving problems AI created and becoming an inefficient interfacing layer for things it made functionally obsolete.

Reply on news.ycombinator.com