It seems like wellness testing is currently in vogue well, testing-wide: from autonomous test generation to self-healing scripts, and every piece in between. But cutting through the big words, I really want to talk about the gritty details of the real world.
Now, has any instructor value been introduced into testing for you; has it been much more of a hype so far?
Which concrete use cases have you attempted to put in action that actually worked or simply failed?
The community could foster some very real stories from the real world.
I actually tried to convert some test scripts into another language and toolset, and unfortunately really didn’t get far at all. There were some generic things it did well (recreating some of the structure), but it took a lot of shortcuts and put in placeholders despite giving it multiple, targeted prompts.
I finally got a test working, but it didn’t feel like it was a smooth experience and I really don’t feel like I learned much in the process.
I certainly don’t think it’s all smoke and mirrors, but I do think the hype is generally overblown.
Have a look out for some of Ben Fellows, experiments, analysis and posts on using AI as an accellerator for test automation. He posts quite a bit on linkedin, involves the community quite a lot inviting them to free training session and for input.
For me he takes a very professional testing approach to evaluating the use of AI in his teams daily work.
I have used some of these with some fairly decent results, it has a strong emphasis on automation at the moment and to get the better results relies on having full access to product code. Things like add Test ID’s, create a POM structure and create an initial set up tests they are looking at hours to do rather than days. His posts explain this better than I could here.
A couple of things I am curious about, is that some people are posting that it is already a complete game changer and they could not go back to not using it.
This raises a question to me regarding how much dull repetitive tasks they were actually doing before, for some it may have been a lot of time with this focus rather than a lot of time on critical thinking based activities. It may carry a risk of doing the wrong things just faster and if that’s the case the progress could be a no-brainer yet still that niggling feeling they should be doing more of that critical thinking side of testing in the first place.
Now the way I use the internet has changed, back and forth discussion based quicker to dig deeper into things and this can translate well into research of a testing risk and ideas to approach that risk or quickly build a prototype tool to help with that investigation. This part I would like to explore more and I do not at this point have enough use of this to present a solid argument for it being a game changer.