982: In Case You Missed It in March 2026

In this 'In Case You Missed It' episode, host Jon Krohn curates key insights from recent conversations on the Super Data Science Podcast, highlighting transformative trends in AI, education, and the future of work. The episode opens with a powerful discussion with author Zach Kass about the potential of AI to revolutionize personalized education, drawing parallels between historical one-room schoolhouses and modern AI-driven learning models. Kass emphasizes that education should restore the sanctity of childhood by fostering self-discovery beyond economic incentives, spotlighting innovators like Eva Moskowitz of Success Academy and Mackenzie Price of Alpha School, who are reimagining classrooms through self-guided learning and student autonomy. The conversation then shifts to AI agents and autonomous intelligence, with insights from Kyungyung Cho on how AI systems can explore and plan in both digital and physical spaces using latent dynamics and world models. Chris Fregley shares his shift from manual coding to relying on AI assistants like Codex and Cloud Code, stressing the importance of running development on production-grade hardware and using automated evaluations over line-by-line code reviews. Lin Xiao of Fireworks AI explains the rise of 'autonomous intelligence'—where AI systems continuously adapt to private, enterprise-specific data—ushering in a future of millions of specialized models rather than one universal AGI. Finally, Rohit Chowdhury discusses the decentralization of data and work, where AI enables human-AI collaboration, redefining career paths around clarity of thought and domain expertise rather than traditional coding skills. The episode concludes with a vision of a more fluid, intelligent operating environment where data, AI, and human judgment converge. Key takeaways include: 1) AI can enable truly personalized education by freeing teachers from one-size-fits-all instruction and allowing students to learn at their own pace; 2) The future of AI development lies in autonomous intelligence—continuous, application-specific model adaptation using private enterprise data; 3) AI agents are evolving into junior engineers, making coding interviews obsolete and shifting value toward clear thinking and structured communication; 4) Small language models (SLMs) will be essential in enterprise AI stacks due to privacy, efficiency, and task-specific performance; 5) The future of work is decentralized—AI systems must operate close to data, not in centralized warehouses; 6) Success in the AI era will go to those who can clearly articulate their goals and leverage AI as a collaborator; 7) The next renaissance may be a spiritual awakening, rooted in redefining human purpose beyond economic productivity; 8) Developers must test on production hardware and use automated evals, not manual code reviews, to stay efficient and avoid reward hacking.