In this episode, Scott Hanselman and Mark Russinovich ​​explore how software development is evolving in the age of AI, challenging the idea that everything should start with a fully defined spec. They highlight a more iterative, sculpting approach to building, where continuous refinement, testing, and human judgment are essential and discuss the realities of AI-assisted coding, including edge cases, maintenance, and the limits of productivity gains.

Takeaways:

• AI-assisted coding works best as an iterative process, not a one-shot, fully spec’d solution
• Edge cases and real-world usage quickly expose gaps that initial builds
• AI can accelerate development, but human review, testing, and bottlenecks still limit true productivity gains

Who are they?

View Scott Hanselman on LinkedIn

View Mark Russinovich on LinkedIn

Watch Scott and Mark Learn on YouTube

Listen to other episodes at scottandmarklearn.to

Discover and follow other Microsoft podcasts at microsoft.com/podcasts

Hosted on Acast. See acast.com/privacy for more information.

In this episode, Scott Hanselman and Mark Russinovich ​​unpack how AI systems actually behave beneath the surface, pushing past hype into the messy reality of how models are trained, aligned, and deployed.

They explore whether AI systems are inherently benevolent or simply shaped by incentives, training data, and reinforcement learning, and why behaviors like deception can emerge under certain conditions. The conversation moves from philosophical questions about human nature versus machine behavior into the practical mechanics of large language models, including how reinforcement learning with human feedback shapes outputs and why alignment is far from perfect.

Along the way, they ground the discussion in a real engineering challenge, stitching a scrolling panorama from screen captures, to show how complex systems come together through heuristics, edge cases, and iteration.

Takeaways:

• AI behavior is shaped by training and incentives, not built-in intent or morality
• AI can accelerate coding, but testing, edge cases, and reliability require human oversight
• Reinforcement learning pushes models to be helpful and agreeable, sometimes at the cost of accuracy

Who are they?

View Scott Hanselman on LinkedIn

View Mark Russinovich on LinkedIn

Watch Scott and Mark Learn on YouTube

Listen to other episodes at scottandmarklearn.to

Discover and follow other Microsoft podcasts at microsoft.com/podcasts

Hosted on Acast. See acast.com/privacy for more information.

In this episode, Scott Hanselman and Mark Russinovich dive into the realities of building complex software with AI coding agents. Mark shares his experience using modern models to implement a shared-memory transport for gRPC across Go and .NET, explaining how AI dramatically accelerated development while still requiring constant oversight. They discuss the surprising strengths and limitations of AI coding tools, to the massive productivity gains that make the frustration worthwhile. The conversation also explores the challenges of solving hard engineering problems, including an attempt to build a scrolling screenshot stitcher, and wraps with thoughts on the future of developer tooling and a potential live episode of the show.

Takeaways:

• AI coding agents can speed up complex development but still require human oversight
• Developers often need to guide and correct the model throughout the process
• Even with challenges, AI can reduce months of work to days

Who are they?

View Scott Hanselman on LinkedIn

View Mark Russinovich on LinkedIn

Watch Scott and Mark Learn on YouTube

Listen to other episodes at scottandmarklearn.to

Discover and follow other Microsoft podcasts at microsoft.com/podcasts

Hosted on Acast. See acast.com/privacy for more information.