Why Do Better Tools Make Me Worse at My Job? (w/ David Epstein) | Monday Advice

What It Covers

Cal Newport and David Epstein apply Eliyahu Goldratt's Theory of Constraints — originally developed for industrial manufacturing in his 1984 business novel *The Goal*, which sold 10 million copies — to explain why digital productivity tools often increase busyness without increasing output, then extend the framework to individual knowledge work and writing processes.

Key Questions Answered

• •Theory of Constraints core principle: Every system is limited by a single bottleneck — the slowest step in any process. Speeding up steps before that bottleneck only creates pile-up, not more output. Goldratt condensed the entire 1,200-page Theory of Constraints handbook into one word: focus. Identify the single constraint first, then direct all improvement energy there before touching anything else in the workflow.
• •Digital tools and the bottleneck trap: Productivity tools like Slack and Notion accelerate information movement efficiently, but that is rarely the actual bottleneck. Epstein's video production workflow illustrates this: scripts moved faster through Notion, but approval still required his personal review. Adding a fact-checker — targeting the real constraint — did more than any workflow software improvement could achieve.
• •Stop starting, start finishing: A 15-person design office making custom gearboxes was switching tasks over 50 times daily. Implementing one rule — no new design starts until a current one finishes — tripled design output within months and cut total gearbox production time from 12 months to 2 months. Limiting work-in-process directly expands throughput at the actual constraint.
• •Identify your personal bottleneck before adding volume: Epstein as a college 800-meter runner cut weekly mileage from 85 to 35 miles and dropped one workout per week because recovery — not training volume — was his limiting factor. Performance improved immediately. Swimmer Sheila Taormina applied the same logic: targeting strength and power rather than aerobic endurance dropped her 200-meter freestyle time by 3.1 seconds, earning her a 1996 Olympic gold medal.
• •Front-load constraint-heavy thinking to accelerate execution: Epstein spent the first year of his *Inside the Box* writing process exclusively on research and mapping, producing a 100,000-word master thought list before writing a single page. A two-day silent retreat at a Franciscan monastery produced a single-page outline. This approach — uncomfortable and seemingly slow — resulted in finishing the manuscript weeks early with no chapters cut.
• •AI tools accelerate non-bottleneck steps: Current AI coding tools create a classic constraint problem: faster generation leads developers to run 13 simultaneous agentic processes, causing log-jams where previously they managed one or two. For academic researchers, AI speeds up plot generation, but data access — often months of relationship-building — remains the actual bottleneck. Faster plots do not produce more papers; targeting the real constraint does.