Systems | Development | Analytics | API | Testing

Recently someone on our team opened a traffic snapshot and found an HTTP request that was captured with our eBPF capture agent, nettap. Our protocol dissector parsed most of the response correctly, but that correctness ended once the response headers were processed. What they ended up with was a recording of an HTTP request/response where the response body was just an incorrect collection of garbage binary data when it should have been JSON text.

I spent the day at NVIDIA thinking I’d be learning about all the newest models and RAM throughput achievements. Instead, a version of the future started to crystallize where AI really is good enough, pervasive and cheap(er). Let me share some of my observations and conclusions.

Every new signup posts a message to our Slack. The format is dull and reliable: Overnight this week one arrived like this: That trailing nothing was the entire incident. No error logs, no alerts. A returning user had signed up, our signup service had attached them to a tenant we deprovisioned back in December, and the only symptom in the whole company was a Slack message that ran out of words.

Ask an AI for K-pop concert advice without saying the group, city, date, or budget. It may confidently send you to a BLACKPINK tribute night in Cleveland with a $400 resale ticket. The AI was plenty confident. It just had nothing real to go on. That is exactly what happens when developers test AI applications with invented traffic. The test may look reasonable. The result may even pass.