Industry executives and experts share their predictions for 2024. Read them in this 16th annual VMblog.com series exclusive.
Generative AI Rewrites the Software Development Lifecycle
By David Brooks, Senior Vice President of Evangelism at Copado
This prediction seems like a no-brainer. Microsoft recently announced they have surpassed one million paying customers for GitHub Copilot. So the use of Copilot for Professional Coders is already well underway, but the real question is how Gen AI will change the entire life cycle of how we develop and deliver code.
Impact of Copilot on Pro Coders
The first impact will not be the sudden reduction in the number of developers. On the contrary, the impact will be on the efficiency and productivity of current coders in the near term. Developers will write less code themselves and perform code reviews for Pull Requests generated by their Copilot. They will become more architect than coder, guiding the Copilot to a solution that conforms to the best architectural design.
Impact of Copilot on Low Coders
The biggest impact on the productivity of enterprise development teams will be the empowering of point and click admins to release solutions that include code. “No coders” will become “low coders” and low coders will be able to do much more by guiding a copilot to build the solution. Pro Coders may still need to perform code reviews and weigh in on the architecture, but Copilots will essentially empower business users to handle more of the actual solution build.
Impact of Copilot on Testers
One big challenge in the development world today is the challenge of creating tests. The Subject Matter Experts (SMEs) who understand the key requirements of new features and business processes are not typically able to create Selenium tests. And those testers who know how to write Selenium scripts do not typically understand the business process in enough detail to write thorough tests. Copilots are not just for writing the application code, they can generate test scripts today. This means that business users will be able to generate test scripts for their new features and business processes without knowing how to write scripts and before the code is even written, which enables Test Driven Development. This will result in increased test coverage and even the automation of repairing existing scripts that get broken when the system is updated.
Impact of Copilot on BAs and PMs
In order to write good code and test scripts, the Copilot must have a clear understanding of the requirements. Product Managers and Business Analysts will have to up their game at writing User Stories and include acceptance criteria. More time will be spent on thinking through the impact of changes before the Copilot is unleashed. BAs and PMs will learn how to use Gen AI to help ensure their requirements are well defined which should enable them to spend more time with actual users to get more clarity on user needs and business priorities.
Agile Sprints as we know them will likely go away. Today we get into a cycle of grooming stories, estimating stories, developing and testing over a 2 or 3 week sprint, then releasing a batch of finished work into production. The future will be more of a continuous development and delivery cycle paced by the refinement of the requirements. It may take longer to research and write the User Story than it does to build and test it.
We will likely see teams meeting in the morning to discuss what they will build and release that day. The architect and UX designer will get the requirements from the BA, discuss their approach, and use GenAI tools to rapidly create working prototypes with the proper architecture. The team will meet again after lunch and review the working product in a Test environment, make any last minute changes, then when they are satisfied with the work, push the big green button that rolls it out to production.
This continuous delivery may work for consumer apps, but in the enterprise we may still see a slower release cycle. Business users do not like surprises. Changes to their business processes must be communicated beforehand and the users must be trained on how the new processes work. If the development teams have more time to design better user experiences and can spend more time on user facing documentation, the biggest impact from GenAI will be the increased satisfaction and efficiency of business users. And that will be a brighter future indeed.
##
ABOUT THE AUTHOR
A 35-year veteran of Silicon Valley, David Brooks is Senior Vice President of Evangelism at Copado. He created the product management practice and led the product team for four years. Before Copado, Brooks spent nine years at Salesforce where he was tasked with launching the AppExchange and led teams that built the Force.com platform.
