Cal Newport unpacks his framework for Slow Productivity, built on three core principles: doing fewer things, working at a natural pace, and obsessing over quality. He introduces "pseudo productivity"—the toxic heuristic that emerged in mid-20th century knowledge work when visible activity became a proxy for useful effort because traditional productivity metrics (Model Ts per hour, bushels per acre) no longer applied. Newport argues that pseudo productivity was tolerable until the digital office revolution—email, Slack, mobile computing—enabled visible activity to be demonstrated at incredibly high frequency, anywhere, anytime, creating a performance theater that drains actual productive capacity. The conversation explores how to build custom AI systems for daily planning (using GPT models trained on transcripts and book notes), the three levels of working with large language models (training from scratch, fine-tuning, and software intermediaries), and why specialized vertical AI will dominate the next wave of innovation. Newport makes the case for abandoning industrial-era proxies and reclaiming knowledge work as a craft that requires depth, patience, and quality over constant performative busyness.

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