Software Engineering in the Age of Creative AIs

URL copied!

ChatGPT exploded onto the scene in November 2022, and there’s no turning back now. Though certainly not the first conversational AI in the market, ChatGPT has opened a lot of eyes to the myriad ways this technology will change the way we live and work.

Transformational change is no longer a far-off, future state concept. In our own lifetimes and careers, says Dr. Jim Walsh, CTO at GlobalLogic, we’ll begin working alongside a truly creative AI – one that not only executes tasks but understands the full context and business needs of the system it’s developing. One capable of developing the required code for a complex system based solely on human-readable user stories, and even running through sprints and iterations automatically, in a constant state of self-improvement, aided and guided by human developers.

Now, Dr. Walsh can’t put an exact timeline on this, of course. But as we’re already seeing AI-aided coding happen today with tools like ChatGPT and GitHub Copilot, he says he expects that in the not very distant future we’ll start seeing AIs that can write more sophisticated software systems based on natural language specifications contained in user stories, combined with formal logical and physical architecture descriptions.

How Creative AIs Will Transform Software Engineering

Can you imagine? In this ebook, that’s exactly what Dr. Walsh has tried to do. In it, he envisions a typical day in the life of a software engineer working alongside a creative AI with the ability to understand and adapt to business context, personas and user roles, and end user needs.

Picture a creative AI with an actionable awareness of the system’s logical and physical architecture, with a well-developed code generating ability and the capacity to deploy code to a testing environment almost instantly.

What might a day in the life of that software development team look like?

In ‘Software Engineering in the Age of Creative AIs,’ Dr. Walsh walks us through this thought experiment together:

What does a future with no coding or coders look like in practice?
Even if creative AIs can generate code that conforms to the specs as given, and even if that code passes exhaustive (and partially computer-generated) acceptance tests enumerated in the user stories, will that be enough to create a software system end users love?
What will the future equivalents of today’s product owners, BAs, architects, engineers and testers do? Will their job consist of writing specs and evaluating the output of the AI against what the product owner and technical architect really wanted, both functionally and technically?
What will purchased software look like in the future?
What will happen when we achieve a constant state of Agile/SAFe sprint review, backlog grooming, product increment planning and sprint kickoff mode powered by creative AIs?
How will creative AI as part of the development team impact the way we develop and package software solutions?

Take a walk through the possibilities and envision a future in which an AI can develop all required code for a complex system starting solely from human-readable “user stories.”

Sure, it’s speculative. But so were Dr. Walsh’s predictions on software and AI from 2009, and over a decade later we know for a fact he wasn’t far off the mark!

Join us in this exploration of how advanced creative AIs may impact our roles as software engineers – and our industry as a whole. We also take a deeper dive into how these future AIs may become true partners in the development process, interactively helping us remove ambiguity from the user stories, identifying missing features and tests, and becoming a full member of the development team.

Click to read Software Engineering in the Age of Creative AIs

More helpful resources: