If you've ever studied computer science in college or taken an online course on programming, there's a good chance you've been exposed to the concept of unit testing, which is creating tests to check that a small portion of your code behaves as you intended.
According to forecasts by analyst firm IDC, by 2028, generative artificial intelligence (GenAI) will largely take over this aspect of coding, and 70% of such tests will be created by artificial intelligence. Some programmers feel that unit tests are also a bit confusing, so having AI write better tests should improve the overall quality of code delivered to end users.
Dhiraj Badgujar, senior research manager for digital innovation, xOps and developer strategies at IDC Asia Pacific, said of the adoption of GenAI to create tests in Asia:
"The Asia-Pacific market for generative AI-assisted software development, including code generation, user interface, testing and other use cases, is growing rapidly. China, India and Japan are investing heavily to improve software testing capabilities. The robustness of the IT industry and emphasis on quality assurance are driving market progress."
With better, more comprehensive unit tests written by GenAI, developers can free up their time to focus on creating new features for the software they are developing. As mentioned earlier, these comprehensive unit tests also catch more program errors, resulting in fewer software errors.
The idea that artificial intelligence will make our software safer is actually very interesting because Rust has built-in memory safety features and is becoming more and more popular.
One of the reasons why software created using low-level programming languages has so many vulnerabilities is because of memory issues. If more software is made in Rust and better unit tested using artificial intelligence, we could potentially see very reliable software by the end of the century.