Automated Testers Advocate for ChatGPT (AI) is like taxi drivers Singing for Driverless cars

It is interesting to analyse the reasoning behind it. By the way, Real Test Automation Engineers won’t believe ChatGPT will ever work for real test automation.

Image Credit: https://www.vox.com/2016/4/21/11447838/self-driving-cars-challenges-obstacles https://www.istockphoto.com/vector/driver-man-pop-confused-gm1314725104-402876759 http://clipart-library.com/clipart/happy-person-clipart_8.htm

https://www.lambdatest.com/blog/using-chatgpt-for-test-automation/

https://saucelabs.com/blog/chatgpt-automated-testing-conversation-to-code

https://club.ministryoftesting.com/t/how-will-chatgpt-change-the-testing-industry/64010

ChatGPT is hot and became well-known quickly (starting December 2022). I expressed my opinion quite clearly in this article: “ChatGPT is Useless for Real Test Automation”. Today, a thought came to me, why do “automated testers” seem excited about this AI Testing? Wouldn’t this will take their jobs away?

I haven't seen/met any Taxi Drivers or Uber/Limousine Drivers who were excited about autonomous cars, i.e. driverless cars. The reason is very logical: if driverless cars are everywhere, they are unnecessary!

Some “fake automated testers” might argue: “I can ask ChatGPT ‘right’ questions?”. No, in end-to-end (UI) testing, business analysts and manual testers probably ask better questions, if ChatGPT provides flawless solutions (manual testers and BAs can copy the script and run it). If ChatGPT cannot provide a properly working script, there is a need to fix broken tests. Anyone with software engineering experience will know it would cost more effort (than hand-crafting).

Consider this fact: the application change constantly and frequently.

Are these “automated testers” retarded to promote something pushing them out of jobs? No, they are just fake “automated testers”. They don’t understand test automation at all, a typical way they live on faking test automation is by jumping to different test automation frameworks/tools every now and then (specifically, after the failure becomes apparent). Something new with a certain level of hype is that all matters.

Does this sound familiar? A software company has tried PhantomJS, WebDriverIO, Protractor, TestCafe, Puppeteer, Cypress and Playwright, within five or six years. Of course, every test automation attempt (with the above) failed, but some “automated testers” remain. They are failure-tolerant. I have met many of these ‘fake automated testers’, the №1 excuse is “there is a new and better test automation tool: X”.