Anyone else drowning in ai-generated noise at work
My team started using ai tools for QA recently. Idea was to catch bugs faster.
It worked for maybe three weeks.
Now I spend more time sorting through garbage reports than I ever spent finding bugs manually. Half the stuff flagged isn't even a real issue, its just the model hallucinating edge cases that would never happen in production. The other half is duplicates of things we already know about, phrased slightly differently each time so they don't get caught by dedup filters. I sat through a wednesday standup last month where we spent forty minutes discussing which ai-generated tickets were worth keeping. Forty minutes. For tickets nobody wrote.
The frustrating part is I can't even say the tools are useless. They do catch real stuff occasionally. But the signal to noise ratio has gotten so bad that I'm starting to wonder if we were more productive before.
Feels like we automated ourselves into more work somehow.