I was looking at the latest automation tools in the market and I see many record&play tool like Selenium IDE that I discovered 4 years ago but now they added this stuff related to AI/ML supposed to never break the execution thanks to solid locators.
I will not mention the name of the tools because I’m not here to do some advertising.
Do you think that it could be useful or in reality we just don’t need AI to fix locators when there are some change in the html code ? Do you think that it could help ? Or it’s just marketing ?
On internet I just see articles saying “AI is the future” but that’s all.
This is just my opinion, but, to me all that is just marketing/propaganda. The harsh truth is that simply none of these fancy expensive tools are as good as they try to make them seem.
I tried a few of them and I prefer custom coded solution - the best way to go if you know what you’re doing, what you want to achieve and have the necessary support, like enough budget.
" none of these fancy expensive tools are as good as they try to make them seem."
oh yes that’s true, but I think that we can use the same concept to any field, no just testing
But as engineers in QA I really wonder if in real life we can be helped by AI etc for our automated tests (I mean : are they trying to solve a real problem ?) . Maybe for apps where ID are dynamic or if the locators are too hard define, I don’t know.
I agree with @mirza , UI automation is still UI automation and will still be flaky, we had some exploration nights with some self-healing scripts but yea… flaky is still flaky
