Is There a Right Rhythm of Work in Tech?

Returning to any conversation about tech culture means sooner or later returning to the same secular religion: hours. Who is shipping late into the night, who is online at 1:14 a.m., who treats sleep as a character flaw, and who, by implication, wants it badly enough.

This is usually where someone invokes 996, the now-infamous “9 a.m. to 9 p.m., six days a week” model that became shorthand for a certain style of high-growth martyrdom in Chinese tech. In August 2021, China’s Supreme People’s Court and Ministry of Human Resources and Social Security publicly moved against the practice, making clear that such overtime arrangements violated existing labour rules.

Tech is one of the few industries that still sells the fantasy that effort compounds visibly. Put in enough hours and you build something, scale it, own it, bend a market around it, and nourish that most modern of ambitions: the hope that your exhaustion is not merely labour, but destiny…

The trouble is that the evidence for this worldview is much weaker than its believers seem to think. John Pencavel’s research on working hours and output found a nonlinear relationship: below a certain threshold, output rises with hours; above it, returns diminish sharply. More time can produce more work, until eventually it mostly produces more time. Schade.

This is awkward for any culture built on visible sacrifice, because long hours are aesthetically convincing. They look serious. Forty-eight-hour hackathons photograph well. They create the comforting impression that effort and value are neatly correlated. But in knowledge work, that relationship is much messier. A tired software engineer does not usually produce a dramatic factory accident, even with unlimited Red Bull and free snacks. They produce sloppier judgment, weaker prioritisation, avoidable bugs, performative responsiveness, and the quiet substitution of presence for usefulness. Research on burnout in software engineering has repeatedly linked overload and high job demands to exhaustion, reduced wellbeing, and ultimately turnover.

Then there is the less fashionable point.. the body keeps score even when the cap table does not. A mildly inconvenient detail. The WHO and ILO estimated that long working hours were linked to 745,000 deaths from stroke and ischemic heart disease in 2016, and found sufficient evidence that working 55 hours or more per week carries higher risk than working 35 to 40 hours. That does not mean every intense stretch of work is pathological. It does mean that “grind culture” sounds considerably less glamorous once translated into epidemiology.

Tech is particularly susceptible to this confusion as its most valuable output is often the hardest to measure. Good judgment rarely announces itself, and neither does careful thinking.

I do not believe there is a single right rhythm of work in tech, but I can certainly see one that feels wrong - a rhythm that treats sustained overextension as proof of seriousness, motivation, and conviction. Do not get me wrong; one must of course work hard and give their best to get what one wants in life. In tech, sprints are real, crunches happen, early-stage companies are not monasteries, and neither are some venture capital funds. But a permanent state of urgency is not ambition.

One could argue, of course, that shorter hours sound suspiciously like a luxury belief for salaried knowledge workers with ergonomic chairs and opinions about boundaries. Fair enough. But even the newer literature on reduced working time is awkwardly unsympathetic to the cult of exhaustion. A 2025 Nature Human Behaviour study, covered by Nature, found that large-scale four-day-week trials improved employee wellbeing while workers reported feeling just as productive. Eurofound’s 2024 literature review reached a similarly inconvenient conclusion: where working time is reduced intelligently, organisations often report gains in wellbeing, motivation, retention, and in some cases productivity itself.

That does not mean every startup should shut the laptop on Thursday evening and wander into a Scandinavian lake, dreamy though that sounds… It means the more interesting question is not how many hours great teams work, but what cadence lets them keep thinking clearly long enough to do great work more than once, while building category-defining companies and genuine venture outcomes. In that sense, I could not resist borrowing an old French phrase: “un esprit sain dans un corps sain”. I fear that in our tech bubble, we like to imagine the mind as infinitely extractable, as if cognition were detachable from sleep, stress, and physiology. It is not.

If there is a right rhythm of work in tech, it is probably knowing when intensity sharpens performance and when it merely flatters the ego. I tend to believe the best teams are not the ones that confuse exhaustion with excellence, but the ones that understand ambition also needs oxygen, and build accordingly.

Painting: Philip Galle after Maerten van Heemskerck’s “The Diligent Worker Aspiring to the Rightousness of the Lord” (1572)

Sources

• Eurofound: Working time reduction with a focus on the four-day week: Literature review

• Graziotin et al: Burnout in software engineering: A systematic mapping study

• John Pencavel: The Productivity of Working Hours

• Nature: Biggest trial of four-day work week finds workers are happier and feel just as productive

• Taipei Times: Coverage of China’s move against 996 overtime practices

• Trends in Cognitive Sciences: How can a 4-day working week increase wellbeing at no cost to performance?

• WHO / ILO: Long working hours increasing deaths from heart disease and stroke