Software Testing Can Be Automated, But Is This The End Of Coding As We Know It?
One of the most repeated statements about artificial intelligence (AI) in the technology industry is that it is about to replace coding as a specialist activity. It’s certainly true that AI has developed very rapidly in the past 18 months and it will change how code is developed, but there is some low-hanging fruit that can be addressed before worrying about the coding process.
Supporting a coding team can dramatically improve their productivity. It allows the programmers to focus on the coding processes by freeing them from some of the manual and essential tasks they need to undertake. This makes them more productive because it allows them to focus, but giving them this focus also improves the job itself, which further aids productivity.
As an example, look at the testing process. Every programmer knows that a small change to the code base can have unforeseen effects. You can’t change one section of code and just test that section only – before the client sees the system, you have to check it all again.
So the introduction of automated testing, in particular smart automated testing that can use AI to vary the functions being tested, removes the drudge work that every programmer hates. They focus on the creative side of their job and allow the AI to ensure that the code works each time a new release is packaged up.
This is true for any form of coding, but it is particularly applicable to mainframe code, where a very large code base and set of testing requirements can lead to corners being cut when testing is only performed manually. The IBA team was in Dresden recently talking about this at the GSE conference.
The IBA Head of Automation Testing, Julia Mikitjuk, recently posted an informative slideshow on LinkedIn focused on the myths around testing and automation. You can go to see the entire slideshow here, but a summary of the key points Julia made are:
- You don’t need to 100% automate everything for an initiative to be valuable
- If you need to invest, then contrast the amount to what will be saved in the future
- Automation can’t be switched on overnight — it needs planning
- Automation might not replace your human testers — the aim is usually to augment what humans can do, rather than to replace them entirely
- Automation helps a lot, but it can’t replace every single aspect of nuanced user interface testing
Software development is evolving. Tools like AI are making a dramatic difference, and many repetitive processes, such as testing, can be largely automated.
However, it is worth exercising caution.
As Julia outlines in her presentation, this is not the kind of project where you can push a button and suddenly it works. Even where a comprehensive testing automation system is in place, there will always be a need to run final checks with people — especially on the user interface.
Coding is evolving and AI offers a very powerful tool that can really help to improve quality and productivity at the same time, but managers responsible for software development should not believe the media hype that coding is finished – even testing code can’t be completely automated.
Automation and similar tools offer a way of reducing repetitive actions, improving quality, reducing errors, and therefore increasing productivity, but they are not an automatic replacement for the software development process. Not yet anyway!
For examples of IBA client success stories, including testing and automation, please click here. Follow IBA Group on LinkedIn for regular updates and comment. For more information on technology strategy and how tech connects to real business solutions please click here.