Vibe Coding: How AI is Transforming the MBA Experience at Yale

As a second-year MBA and co-president of the Yale SOM AI Association, I’ve witnessed a remarkable transformation in how business students approach technology. Last winter, I gathered 40 fellow MBA students and Professor Kyle Jensen to form a learning group called “Coding with Kyle” (CWK). This initiative was born from a simple realization: if I, a former civil engineer with no formal computer science education, could build AI applications, so could others.

My coding journey started in the fall of 2024 when I took CS50, a foundational Yale College computer science course, alongside several AI-focused SOM courses, including Management of Software Development and Large Language Models. The contrast was stark; while traditional coding felt daunting (I still have flashbacks to programming coursework from my freshman year in college), AI coding opened new possibilities. After building my “AI Digital Twin” (a conversational AI version of myself trained on my journal entries) and “AI Journal Analyzer” (a tool that processes 375,000 words from six years of personal journal entries to identify emotional patterns and growth trends), I recognized an opportunity to scale up my work and take others along my AI coding journey.

That first CWK session in December was a virtual Zoom room filled with classmates who had never written a line of code but were eager to learn. As Professor Jensen, a few technically-inclined students, and I demonstrated how AI tools could transform a simple prompt into functioning code within minutes, I watched faces shift from skepticism to amazement. Professor Jensen masterfully created a safe space for experimentation, encouraging us to share our successes and inevitable stumbles as we began applying these tools to everything from personal projects to startup ideas. We’ve encountered the limitations of current AI tools—particularly when building complex systems or debugging subtle errors where the AI sometimes confidently suggests incorrect solutions. These moments remind us that human judgment remains essential, even as the technical barriers fall.

What we’re practicing might best be described as “vibe coding,” a term coined by former OpenAI researcher Andrej Karpathy that has gone viral in tech circles. Vibe coding, or coding enabled by AI, isn’t about mastering syntax or algorithms. It’s about surrendering to the flow—as journalist Kevin puts it, building software by “describing a problem in a sentence or two, then watching a powerful AI model go to work building a custom tool to solve it.”

Traditional large language models like Perplexity, Claude, and ChatGPT, combined with AI coding tools like GitHub Copilot, Bolt, and Cursor, have democratized software creation. As the Hard Fork podcast noted, people like me who lack formal coding experience are now able to build their own programs from scratch. The barrier is no longer knowing how to code; it’s knowing what to build and why it matters for you.

What started as a winter break experiment has since become deeply meaningful to me. Twenty of us still meet monthly at the Program on Entrepreneurship suite in Evans Hall to share projects, grab lunch, and learn from each other. Our group of regular attendees includes former computer science engineers, current startup founders, and plenty of finance and consulting professionals who never wrote a line of code before this project.

Our community has produced creative projects that showcase what’s possible with AI-assisted development. Prasad Fadke ‘25 transformed unstructured Slack conversations into a searchable course review database, while Jaryd Raizon ‘25 developed Boola, a tool that automates venture research and market analysis. Even seasoned professionals have found value in the program. Yeemon Thant ’25, an experienced product manager turned entrepreneur, told me that “AI tools have compressed my prototyping cycle from weeks to hours, dramatically accelerating both my creative process and execution.”

For me, CWK represents more than just learning to code. It’s about creating a space where MBAs can experiment without judgment; where we acknowledge that traditional career paths and problem-solving methods are evolving; and where we support each other through the frustrations and victories of building something from scratch. I’ve found immense joy in watching classmates light up when they create their first working application.

For any MBA student reading this: you don’t need to wait for permission or formal training to start your AI journey. Find your community, embrace the AI tools available, and start building. The line between “technical” and “non-technical” professional work is blurring, and I’m proud that Yale SOM is at the forefront of this change. Coding and prototyping with AI can initially feel strange, but those who give it a try can unlock something important: confidence in bringing any idea to life.