Lately my AI agent is able to implement a feature within 30 mins or so but I keep asking it to review the changes again and again until it finds no issues. And, this is taking me a full day (or two) to release that feature.
It’s not fast but def avoids serious issues from
If I deployed that to production without a code review, my database would be in outage and my API would have downtime.
Agents write a few lines of code that run queries against the database in a loop on a hot path because they don't "see" and "understand" what they do.
Of
I love how many people are attaching to the "half-bounch" startup time metric. I used to do paid macOS dev ~20 years ago and most macOS devs I knew then used the bounce metric! Back then multiple bounces was the norm due to HDDs. One bounce really wasn't possible until SSDs.