What will the next generation of developers do when AI writes the code

When AI can produce code in seconds, what is the job of a software engineer? For someone graduating with a computer science degree, what is the future of your daily work if it's not coding? And when AI is designing the architecture, writing the code, and fixing the problems, how does a software engineer learn software engineering?

Microsoft and Carnegie Mellon University are betting that the career of a software developer is here to stay-and that the junior developers we train today will be the senior engineers of tomorrow.

That's why we launched PRAISE: an internal preceptorship program that pairs junior engineers early in their careers with senior engineering preceptors. PRAISE stands for Preceptorship for AI-accelerated Software Engineering.

We borrow the preceptorship concept from nursing, where new nurses are paired with experienced nurses to build real-world experience and confidence. The goal isn't to keep newcomers coding as though AI never arrived. It is to help them develop the understanding, confidence, and accountability to build in the new ways we're all learning in real-time.

The partnership links practical engineering experience with academic instruction, rather than asking either academic institutions or companies to solely carry the responsibility of training the next generation of engineers.

Does a CS degree matter anymore? Yes, it does

People asking if software engineering degrees still matter when AI can code-and any human can use prompts to build an app-are asking the wrong question. Would these same people ship AI code in a finished product sight unseen (please, we implore you, say no)?

Of course we need junior software engineers. Senior engineers ensure the technology we ship works safely and effectively, and today's junior engineers will be tomorrow's senior engineers.

Coding with AI is only one part of the evolving software development lifecycle. More of our time is now spent on design and architecture than coding every line. But that doesn't change the need for us to understand what the underlying code does, even if we're not writing all of it by hand. After all, when a car breaks down, it takes someone who knows what's going on under the hood to get it back on the road.

With PRAISE, preceptors coach their preceptees to use AI tools thoughtfully, while building strong systems thinking, taste, and judgement. Just as we evolved and abstracted certain elements of software engineering by using compilers, engineers are now adapting to using AI where it provides efficiency in the overall process of creating a product.

AI changes the work; it does not remove human responsibility for outcomes. Preparing the next generation of engineers means giving them access to both AI tools and experienced guides, so they can grow from producing code to deciding what deserves to be built in the first place.

How the preceptor and preceptee model works

What does the PRAISE model look like in the workplace? Similarly to the nursing model of preceptorship, the key is creating space for junior engineers to workshop questions and problems with senior engineers. Senior engineers need to understand that their junior team members might not always have the confidence to contribute to projects at the speed and scale possible after years in the industry. Building that confidence and rapport with their preceptor gives junior engineers room to learn by doing-not simply watching someone else solve problems.

The preceptor also shares their insights, explains their taste and approach to solving a problem, and provides tips with examples. They can also advise the preceptees on where to focus based on their skills and knowledge.

We teach junior engineers to test assumptions and discuss unfamiliar changes with their colleagues. Mentorship helps turn the process of producing a solution into deliberate learning.

That judgement for when to use AI should be viewed as ongoing education, which is why we're partnering with Carnegie Mellon University on curriculum as part of the PRAISE program. The intent is to set students up for success in the workplace and accelerate preceptees with formal curricula to complement the guidance from the preceptor on how to build and deliver with confidence.

The courses give them early experiences in how to deploy production-ready solutions. Preceptees will take classes and participate in small group discussions to bridge theoretical concepts of computer science to real-world software engineering practices and their work at Microsoft. The curriculum at Carnegie Mellon University augments the on-the-job learning that engineers get at Microsoft from their teams and preceptors, adding formal classroom education to their training.

"It used to be that early-career software engineers would do a lot of tasks that would build up their expertise," said Michael Hilton, Teaching Professor & Associate Department Head for Education in the Software and Societal Systems Department at Carnegie Mellon University's Kenneth C. Griffin School of Computer Science. "Combining forces between academics and practitioners allows us to bring our expertise together and create solutions that draw on computer science research while standing up to the daily realities of software development."

We're excited to work with Carnegie Mellon University to train the next generation of software engineers, helping to maintain a healthy talent pipeline for the future workforce.