AI and Test Automation

I see lots of different articles, solutions etc about artificial intelligence which makes me think about the future of human being as well as my own proficiency: test automation. The latest one was an interview of Serdar Kuzulo?lu with Yuval Noah Harari which was very inspiring but in the same time gave me a whole new idea about the relationship between AI and test automation.

Most of the sources and solutions I've seen so far about AI in test automation focuses on creating test scripts. I doubt that this is the best way to use AI in test automation. Test creation is relatively the easiest part of test automation. Many commercial tools have recording functionality (which I don't advise to use other than getting first or sample codes). Other options to easily create test scripts are BDD and keyword driven techniques. I've never had a chance to use an AI generated script so far but I don't think it'll be as readible as a script written by a tester in BDD.

So I believe AI will have better uses for test automation in data generation, failure inspection or forecasting.

What came to my mind when I watched the interview was: AI will not be used in test automation as much as test automation will be used to test AI. One million dollar question is not how to use AI in test automation, but how to test AI with test automation! People will be more interested in testing AI as much as possible, as it is the real unknown feared by humans.

So as AI solutions will come to the stage in many different areas, it will be more and more significant to test them again and again. Humans will try to make sure that AI does what it's supposed to and more importantly it is under control!

To sum up, I believe test automation will help us more in testing AI, than AI help us in test automation.