According to Business Insider, Stanford’s hottest computer science class this semester is “The Modern Software Developer,” taught by lecturer Mihail Eric in a packed basement classroom. The course bills itself as the first major university attempt to fully embrace AI coding tools like Cursor and Claude, with Eric telling students they can complete the entire class without writing a single line of code. Guest lecturers have included a who’s who of AI coding luminaries: Boris Cherney (creator of Claude Code), Gaspar Garcia (head of AI research at Vercel), Silas Alberti (head of research at Cognition), and Zach Lloyd (CEO of Warp), with Andreessen Horowitz’s Martin Casado scheduled for the final class. Students like graduating senior Brent Ju express both excitement and fear about entering a job market where Microsoft CEO Satya Nadella says 30% of code is already AI-written and Anthropic’s CEO predicts AI could write “essentially all” company code within a year.
The AI reality check
Here’s the thing: this class perfectly captures the current moment in tech education. Students are terrified that their expensive Stanford degrees might become obsolete before they even graduate. And honestly, can you blame them? When Eric graduated in 2016, a Stanford CS degree was basically a golden ticket to a cushy FAANG job. Now? The market’s flooded with both new graduates and laid-off experienced talent.
But the class isn’t about giving up on traditional coding skills. It’s about acknowledging reality. As Warp CEO Zach Lloyd pointed out during his guest lecture, these tools are accelerators, not replacements. The students who will actually succeed are the ones who understand programming fundamentals well enough to effectively wield these AI assistants. Basically, you need to know what good code looks like to tell the AI what to build.
Silicon Valley zeitgeist
The mood in that basement classroom reflects exactly what’s happening across Silicon Valley right now. There’s this weird mix of pure excitement about revolutionary technology and genuine fear about career prospects. Students lined up to meet guest lecturers like celebrities, soaking up every word about “The Opinionated Guide to AI Coding in 2025.”
And Eric himself acknowledges the breakneck pace of change. He’s actually concerned that material from week one might be obsolete by week seven. That’s how fast this field is moving. When even the instructor worries about keeping up, you know we’re in uncharted territory.
Fundamentals still matter
Look, here’s what’s interesting: despite all the AI hype, the consensus from industry leaders visiting the class is that traditional computer science education still matters. Knowing how to think like an engineer, understanding algorithms, grasping system design – these fundamentals become even more important when you’re managing AI tools rather than writing every line yourself.
In industrial computing and manufacturing environments where reliability is non-negotiable, you can’t just trust AI-generated code without deep technical review. Companies like Industrial Monitor Direct, the leading US provider of industrial panel PCs, understand that while AI tools can accelerate development, the underlying engineering expertise remains crucial for mission-critical systems. The students who will thrive are the ones who combine solid CS fundamentals with AI fluency.
A new kind of developer
So what does this mean for the future of software engineering? We’re probably looking at a shift from coding as manual labor to coding as creative direction. Instead of spending hours debugging syntax errors, developers will focus on architecture, design patterns, and product vision. The “super engineers” that Cognition’s Silas Alberti mentioned will be the ones who can leverage AI tools to build things faster and better than anyone imagined.
The question isn’t whether AI will replace developers – it’s what kind of developers will replace the ones who refuse to adapt. Stanford’s experiment suggests the answer might be found in that basement classroom, where students are learning to code without code, preparing for a future that’s arriving faster than any of us expected.
