Kevin Surace says developers rarely want to own QA work even when DevOps says they should. AI will automate the repetitive testing, pushing QA teams toward strategy and business communication instead.
You think about. I think one of the critical tension points that as I work with software teams is between the developers and software qa. They're kind of in the game together on the same team, but largely opposing many of the times. What happens to that relationship and do you think that the developer is going to take on or be able to take on actually running scripts and implementing this on their own or do you think guided by the QA team or what happens to that sort of software supply chain as AI steps into the middle of it?
Well, you know, the entire DevOps modality or culture says there shouldn't be any QA that has not happened in, you know, much more of, I don't know, 10, 15 years of DevOps effort. It does happen in places like Facebook or Amazon. There are situations where you can do it, but in general that hasn't happened. And the reason is it's a little bit fox watching the hen house. You know how when a developer gets mad when someone in QA finds a bug and they're mad at the person for finding it? Right. It's just that's nature. That's human nature. Right. It's human nature. And so what happens is if the developers have to find all. Now developers are clearly doing unit testing, by the way. They have to find all their own bugs in an integrated system, an end to end system. You know, it's not, by the way, it's not the best use of developers, it's not the best use of developer time. It's not what they want to do and they really don't want to find the bugs so. And they're not incented to do so. Like none of that works culturally and I know that that's core to the DevOps culture and I've just not seen it in general work well. What works well is to have a high quality QA team that is strategic and I think ultimately not writing scripts and not doing manual testing, they rely on the AI for that, but they're very strategic about we found, you know, in 52 issues. Let's prioritize those issues. Let's think about that priority. Let's talk to the business analyst, let's talk to the product manager. Let's get development involved and understand the ETA on some of these things. All of that becomes important. Let's develop more test data, et cetera. Right. Look, I think it's a better future. I think we're going to get people out of the mundane jobs and repetitive jobs of scripting or manual testing. Yes, those will go away. But that's okay. We're going to automate those portions, and the rest become even more important and more strategic.
