[deathanatos]

My company has 3 AI on every pull request now. They behave as follows:

1. a general coding AI: Completely broken. Should auto-comment, but never does anymore. Stopped a while back, nobody seems to know why.

2. another general AI: You have to at-chat it. It reacts to the message with <eyes emoji>, but never actually posts a comment?

3. a security bot. Comments, when it thinks there's a problem, in the most obtuse way possible. "SAST findings". But the findings are behind a link, and none of us devs are given access.

I could lean on and press the various people shoving AI down my gullet to like … look at this, and the actual lived experience of devs trying to derive productivity from this mess? But IDK what's in it for me, really.

Even Claude, when it worked, would comment in the most sociopathic manner possible: an English prose description of the problem, attached to an utterly unrelated line of code. Part of that is probably Github, who does let you attach comments to arbitrary lines of code in a review, only the blesséd lines can have comments. Literally none of our AI can format their complaint with a freaking suggested change (i.e., the Github feature, no, instead I get English prose).

Honestly for all I know we failed to pay the bill or something inane, but it would be nice if the AI could format an error message, or something.

[economistbob]

I ran Qwen 14b distillation of Deepseek locally once and created a console app and recoiled in horror at how powerful those things are. The problem is that humans are not designed for default deny mental processing, and so I am not optimistic about long range effects since the hallucination aspect can never be eliminated. No one is going to check everything for errors. A single error in some spreadsheet formula or code somewhere can have catastrophic consequences for years compared to what could have been. I see the transition from expert systems to language models to be a travesty foisted upon mankind. Pushing efforts to drive such language model outputs into data wrangling and analysis is especially heinous because that affects humans' very lives and ability to thrive or survive, or even face criminal consequences.