IT Home On October 17, Microsoft posted a blog post yesterday (October 15) announcing a partnership with Carnegie Mellon University to launch the “AI Acceleration Software Engineering Mentorship Program” PRAISE. The goal is to maintain the source of future senior engineers while accelerating AI-driven development.
When AI can generate code, participate in architecture design, and fix bugs in just a few seconds, does there still room for junior programmers to grow? The answer provided by Microsoft and Carnegie Mellon University is: yes, and training methods must be redesigned.
According to the blog post cited by IT Home, PRAISE draws from the “clinical mentoring” model in the healthcare industry: early professional engineers are paired with senior engineers, who guide them in understanding the system, testing hypotheses, deciding when to use AI, and being responsible for the final results. The project emphasizes that the goal is not for novices to “pretend AI doesn’t exist” and continue writing code manually. More importantly, it aims to develop system thinking, engineering taste, judgment, and a sense of responsibility.
Microsoft believes that AI changes the nature of work, but it does not eliminate human responsibility for software quality and safety. As more time is spent on design, architecture, and verification, engineers still need to understand the underlying code logic; otherwise, when the system fails, the team will have difficulty diagnosing and fixing the issues.
Carnegie Mellon University will provide formal courses and group seminars for this project, helping students integrate computer science theory into real production-level software delivery scenarios. Michael Hilton, a teacher at the university, said that in the past, junior engineers built up their professional skills by performing a large number of basic tasks. Now, both academia and industry need to jointly design new growth paths.
