rubber duck vs ai assitant

Embracing ML in Software Testing

Part I: Leveraging Collaborators and ChatGPT Assistance

Sometimes, you may find yourself writing a draft of a particular diagram or a piece of code in an isolated medium to check if it does what you think it does. Alternatively, while at work, you might suddenly wonder, “What if I were to use Selenium for UI testing?” From there, you brainstorm the whys and hows, and so on…

These days, we have valuable assistance from tools like ChatGPT, Google Bart, Bing, and others. How helpful are they? Have they improved my work?

Well, I can fairly say they have. It’s nice to have such tools that can boost our productivity by quickly answering our questions about different methods, and errors, or even proposing solutions to our coding problems.

Remember the rubber ducky method? Well, I think that nowadays, this is quite the equivalent.

Advantages of AI collaborators:

  • You can get quick feedback on your ideas;
  • You can get more creative;
  • Get quick answers to different questions;
  • Improve your proposed solutions;

Disadvantages:

  • You may feel that you’re cheating, like entering the solution from the back of the book;
  • Trusting too much, not everything that AI gives you should be taken as 100% good, sometimes, AI may fabricate answers just because it sounds good, so you need to carefully choose what to enter and what to pick.

Certainly, there are more advantages and disadvantages, but these are mainly the few that come to my mind.

In conclusion, I use and strongly recommend healthily utilizing AI collaborators in a manner to enhance your software testing experience.


Posted

in

by