David Epstein on Constraints, Attention, and Narrative Values

Interview with David Epstein, science journalist and author, conducted by William Green (RWH068). Epstein is the author of The Sports Gene (2013), Range: Why Generalists Triumph in a Specialised World (2019, NYT #1 bestseller), and the new book discussed throughout this episode: Inside the Box: How Constraints Make Us Better (2025).

Key ideas

Constraints unlock rather than stifle potential. Herbert Simon: “It is a myth, widely believed but not less mythical for that, that people are most creative when they are most free.” Limits force prioritisation and launch productive exploration — the thesis running through Epstein’s work from the periodic table to Pixar.
Attention is scarce and degrades through fragmentation. Simon in 1970: “a wealth of information creates a poverty of attention.” Gloria Mark’s 25-year study shows average task focus has collapsed from ~3 minutes to 45 seconds; attentional residue means switching harms all subsequent tasks. The design principle is “attention must be preserved,” not “more information is better.”
Theory of Constraints (Goldratt): the bottleneck determines total output. Output of any system is limited by its single least efficient step. Fixing the bottleneck — not optimising everything else — is the highest-leverage intervention. Applied to knowledge work: for most decision-makers, the bottleneck is their own attention.
Satisficers beat maximisers. Maximisers (who try to evaluate all options and find the best) are measurably less happy with their decisions, more prone to regret, and more likely to prefer reversible solutions — which also make them less happy. Satisficing (Simon’s portmanteau of satisfy and suffice) is both pragmatically and psychologically superior.
Narrative values consolidate caring. Identifying animating themes in your life story (curiosity, open-mindedness, forgiveness) creates a coherent identity and a filter for the infinite things you could care about. A dense network of reciprocal obligations — inconvenient, socially synchronised commitments — is central to human thriving; too much autonomy generates decision fatigue and isolation.

Herbert Simon and the scarcity of attention

Simon (1916–2001) was one of the most important thinkers of the 20th century: co-creator of the first AI demonstration (Turing Award), founder of cognitive psychology, Nobel laureate in economics. His unifying subject across all these fields was human decision-making under limits.

Bounded rationality: humans do not evaluate all options and choose the best (the rational-actor model). Instead, they satisfice — take a good-enough path. Simon argued this should be proactive: wear the same socks, eat the same breakfast, have the same routine. This prevents Fredkin’s paradox, where most time is spent on the least important decisions (because the options are so similar, spending more time makes no difference — yet that’s where we agonise).

Information and attention: Simon’s 1970 lecture formulation: “In an information-rich world, the wealth of information means a dearth of something else — a scarcity of whatever it is that information consumes. What information consumes is rather obvious. It consumes the attention of its recipients. Hence, a wealth of information creates a poverty of attention.” This was 30 years before the internet. The design principle that follows: treat attention as a resource to be preserved, not a pipeline to be filled.

The maximiser problem

Maximisers — people who try to optimise all choices — are reliably less satisfied than satisficers. Evidence:

Less happy with decisions; less happy with their lives
Much more prone to regret
More likely to prefer reversible options (even though reversibility makes them less happy; commitment is what generates satisfaction)
Maximising tendencies appear to be rising — the leading theory is that the internet makes it trivially easy to continue comparing yourself to alternative choices after a decision is made, preventing closure

Infinite scrolling, as a specific mechanism: since its introduction, international surveys show people have become progressively more bored — the existence of something else you could be watching spoils the experience of what you are watching.

Attention fragmentation (Gloria Mark’s research)

Gloria Mark (UC Irvine) has monitored knowledge workers for 25 years:

Late 1990s: average task focus of ~3 minutes before switching
~10 years later: ~1.5 minutes
2022: ~45 seconds (where it has plateaued)

Multitasking is not possible: switching requires stopping one task and activating a new rule-set, leaving attentional residue — cognitive carry-over that occupies working memory and degrades performance on the new task. More toggles = lower productivity + higher measured physiological stress.

Practical interventions:

Batch email into 2–5 blocks rather than 77 individual checks
Schedule monotasking blocks where only one class of work is active
Place the phone outside the room: attention retraining takes only days to show measurable improvement
Cognitive outsourcing: write down intrusive thoughts so they leave working memory

Theory of Constraints (Goldratt)

Eliyahu Goldratt was an Israeli physicist studying crystal atomics when a friend asked him to analyse a chicken-coop-building business. He found work piling up at one step; moving one worker there tripled output. The insight became the Theory of Constraints: total output of any system is determined by its single least efficient step. Optimising every other step cannot increase total output until the bottleneck is addressed.

Goldratt’s summary: “Can I summarise all of Theory of Constraints in one sentence? I think I can do it in one word. Focus.”

Case studies in the episode:

Custom gearbox company: design office (15 people) toggling between projects ~55 times a day was the bottleneck for the entire firm. Rule implemented: “stop starting, start finishing” — no new design until the current one is complete. Result: tripled design throughput, total lead time cut from 12 months to 2 months.
Sheila Taormina: US swimmer failed to make 1992 Olympics in 200m freestyle. Took Management 577 for a final class; applied TOC to her training as a class project. Identified her bottleneck as strength/power (not aerobic capacity, which her coaches had been developing). Changed coaches; focused training on the limiting factor. Four years later: exactly 3 seconds faster; 1996 Olympic gold (relay). Unretired for triathlon → US national champion, two Olympic appearances. Then fencing and horse-jumping → modern pentathlon, two more Olympics. Only woman to compete in three different sports across four Summer Olympics. All of this from one management class in 1992.
Personal bottleneck: Epstein’s own 800m walk-on career; trained down from 80 miles/week to 35 to improve recovery (his bottleneck); became a university record holder and US national competitor.

Applied to knowledge work: for most decision-makers and leaders, the bottleneck is attention. The fix is widening the bottleneck — delegation, batch processing, eliminating unnecessary toggles — not optimising steps that aren’t the constraint.

See Theory of Constraints.

Constraints and creative work: Pixar vs. General Magic

General Magic (early 1990s): created the iPhone 20 years too early. Had the designers of the original Mac, Goldman Sachs concept IPO, every resource imaginable — and no constraints. Every idea someone had was added. The project grew out of control. They defined their customer as “Joe Sixpack” and after years of missed deadlines realised no one knew who that was. Bill Gurley: “More startups die of indigestion than starvation.”

Pixar (same era, parallel trajectory): Ed Catmull’s governing principle was structured creativity. Key practices:

Directors spent years in story development with small teams, stripping away characters to find the core — before entering expensive production. Seemingly inefficient; actually optimal over the whole cycle.
The three pitches rule: directors anchor on their first idea; you must pitch three. You are not allowed to pitch one.
The beautifully shaded penny problem: animators would spend weeks perfecting background detail no audience would notice. Catmull’s fix: velcro popsicle sticks on a board (each = one animator-week). To keep polishing the penny, you had to remove sticks from a character that still needed work. Made trade-offs visible. “Bumpers in the bowling alley.”

Catmull to Epstein: “There are tons of good ideas in organisations. What’s comparatively rare is a structured system for channelling those ideas into something that actually matters.”

Isabelle Allende and structured creativity

Allende did not start writing until nearly 40 (now 80 million books sold). In exile from Chile, she began writing her dying grandfather a letter recounting his stories; it became a novel. Since then, she has started a new book every January 8th — the same date — for 40 consecutive years, producing roughly one bestseller every 18 months.

Her daily practice: light a candle to begin, blow it out to close. The same hours, silence required. First several weeks of each book are throwaway warm-up. Historical research provides the structural frame. When profiles describe her as a “magical medium” who channels characters, they are projecting her magical-realist subjects onto her working method. In reality: extraordinary discipline, ritual, and seasonal structure.

Her memoir about her daughter’s death (Paula) broke the cycle — she thought she was done writing. She skipped one January 8th. The next one came, she sat down anyway, and the process began again.

Narrative values and the consolidation of caring

The philosophical model that closes the book: humans need a small set of narrative values — animating themes in the story of their life. Not abstract aspirations but recognisable through-lines: curiosity, open-mindedness, diligence, forgiveness.

The function: in a world that rains down an infinite number of things to care about, narrative values consolidate caring. They create a coherent identity (linked to well-being) and establish “personal policies” — conceptual constraints that make life manageable and meaningful. As Epstein writes: “This is who I am and this is what I can do.”

The Harvard Study of Adult Development (86 years, multi-generational longitudinal study): the central finding is that a dense network of reciprocal obligations is crucial to thriving. This cuts against pure autonomy-maximisation: you have to inconvenience yourself to sync up with other people’s schedules. Epstein’s experience: after optimising fully for autonomy (independent writer, no fixed schedule), relationships suffered and he was not thriving. He joined a board, started dance classes — reinflicted schedule constraints — and satisfaction returned.

Jonathan Haidt: “It is not good for anyone to have access to everything everywhere all at once.”

Robert Johnson and the creative graveyard

The Delta blues guitarist’s mythology (sold soul to the Devil at the crossroads for musical talent) dissolves under reporting: Johnson and a slightly better guitarist retreated to a rural cemetery most nights to practise, because it was quiet. The real secret to his transformation: “silence and structure” — the same formula Allende describes, available to him only because he left ordinary life and went where no one would interrupt.

Theory of Constraints — Goldratt’s system; bottleneck logic applied to business and individuals
Compounding — constraints as the channel that focuses compounding ability into output
Knightian Uncertainty — adjacent theme of what to do when information is abundant but clarity is not
William Green — interviewer; RWH host