About this episode

Scott and Mark explore the complexities of AI systems, their training, and the implications of their behavior.

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&nbsp…

More episodes of Scott & Mark Learn To...

• Scott & Mark Learn To... ZoomIt, Evolved · May 6, 2026 · 27 min
• Scott & Mark Learn To...Vibe Coding, for Real (Again) · March 25, 2026 · 25 min
• Scott & Mark Learn To...A Public 1-1 for Software Engineering Preceptorship · March 11, 2026 · 15 min
• Scott & Mark Learn To... Are Apps Dead? · February 4, 2026 · 13 min
• Scott & Mark Learn To...Pilot Coding - Vibe Coding for Real · January 21, 2026 · 34 min
• Scott & Mark Learn To... Get Ready for Ignite · January 7, 2026 · 41 min

Explore listener stats, chart rankings, contacts and more on the Scott & Mark Learn To... podcast page.