recently I gave AI a task to configure a software and the software didn’t have any docs of how it works only the source code. the agent went through all the files and generated a summary of how it works and why it’s not working for my particular scenario and suggested edits to my docker-compose.yml file. I couldn’t believe it since the bug was very hard to find and it used headless firefox to find why it wasn’t working.
made me realize do we still need documentation of how a software work when a AI can easily explain it?
The days of constant hallucinations are over.
I get it - I was pretty much where you are a six months ago. AI just produces low quality slop. You have to watch it like a hawk because it will just bullshit some plausible but totally wrong solution. It will write tests that don’t test anything, go round in circles making nonsense changes trying to fix things that are fundamentally wrong.
Yeah, those days are gone. The latest models (at least Astra 6 or Opus 5.5) are actually competent.
I have no skin in the game in convincing you but you’re pretty much in the “the iPhone is only popular because of Apple fanboys” stage of denial. I guess the only real cost for you is wasted time and being late to the party, but… that’s probably still bad.
And no I’m not super happy about this. I mean it’s great that I can now easily vibe code the tools I always wanted - and the vibe coding actually works, unlike 6 months ago.
But it also doesn’t bode well for me having a job and I like having a job. 😔
But I’m not going to stick my head in the sand and pretend that AI is still shit just because it would be great for me if that were the case.