Thanks again to BrowserStack for hosting my talk, “The Death of Test Engineering” at their <break>point 2026 online conference. Everyone I worked with was fantastic and I had a great time giving the talk, answering questions in the chat, and checking out some great speakers from the lineup. Hopefully I rattled the cage a little and inspired some testers to start thinking about their craft in a new light.

As it’s generated a bit of “feedback”, I wanted clear a couple things up – my talk is a critique of the testing industry’s obsession with automation frameworks, script counts, and coverage metrics. Much of what “test engineering” has become is due to a maintenance economy centered on keeping automation systems alive rather than improving quality or managing risk.

I have always believed the hardest problem for testing isn’t the mechanics of test automation; it’s understanding the system well enough to create insights into risk under the constraints of the program. Test engineering is “applied systems thinking” and that kind of work doesn’t disappear because AI can extrude Playwright scripts faster than humans. 

Vendors have made fortunes selling the illusion that tests are testing, and now AI is accelerating the contraction in the number of opportunities for testers due to that entrenched idea. So if someone listens to my talk and only hears “all test engineering is just script creation and execution,” then I’ve either overstated my case or they’ve missed my point. (Probably both!)

Anyway, enjoy!