How AI can make us better developers

It might not reduce our workloads, but it could do something even better.

Worries about technology taking away jobs is not unique to AI in 2023. Similar concerns emerged about the internet in the 1990s, computerization in the 1970s, and even electricity way back in the early 1900s. It can be argued that each of these technologies provided new jobs and economic value – covering many times over the jobs that were lost. It might be too early to say the same of the current AI boom, but there are reasons to be optimistic, and we can look at history for some guidance.

My own opinion is that AI will be more of a partner than a threat if harnessed well. Already, we are seeing artists creating interesting new projects using generation models like MidJourney and GPT-4. There is no reason why we in software and testing should not be doing the same, making these tools part of the development process in a controlled manner. I feel there are three ways AI can be harnessed to help us become better at our craft rather than redundant.

As a pair programming partner

Consider an AI tool as your assistant that works in parallel with developers by analyzing the code and providing suggestions. It might also be useful to develop user interface designs or create sample database tables, and make them available for the next stage of development.

While many are spooked by AI’s ability to generate lines of Python and RoR, it is important to realize that AI doesn’t have the creativity and audience understanding that you do and it should be limited to do programming “gruntwork” only. Similarly, QA professionals can leave the basic checks to AI while focusing on edge cases and those scenarios that need more of a human touch.

As your assistant

Some might say writing code is the easy part and that reviewing it is the tough part. Indeed, figuring out if subject matter experts are needed, notifying architects and leads, and coordinating a time that works for all could lead to delays. That by itself could lead to the project getting sidelined, or missing a deadline to review documents. Thankfully, here’s where AI can act as your secretary. Imagine if it scoured the code (and your input) and figured out which kinds of experts are needed, which architect is available, looked at everyone’s calendar and auto-generated a meeting in a matter of seconds. The productivity gains would be immense – not to mention all the mails and Slacks saved!

As a starting point for the unfamiliar

Many times, we as developers need to work in new industries or geographies where we may not have expertise. Given the current global nature of software development, a request to a client for a reference might waste valuable time. Here’s again where AI can help. Suppose you’re working on a tax module for a financial client and need a tax submission sheet, and a tool would be able to generate one for you based on your needs. Similarly, it could help business teams with knowledge gathering so they can develop a better requirements document.

In summary, I feel that AI will be transformative to the way we do work and that it’s in our best interest to learn how to leverage it. I do not think it will reduce our workload (sorry!), but it will make us better developers and help us achieve our full potential. By addressing the mundane, it will help us focus on what only humans can do.

My advice right now is to be curious and play around with these tools – inside and outside of work.

How AI can make us better developers

As a pair programming partner

As your assistant

As a starting point for the unfamiliar

Author

Agathiyan Athimoolam

Related Blogs

CORPORATE OFFICE

SERVICES

PRODUCTS

Stay in the loop, join our mailing list