Why High Agency Has a Biology (And What to Do on the Days You Don't Have It)

Quick answer: High agency means questioning whether something could be different, then acting to change it. At work, that includes challenging an unhelpful brief or testing an alternative to an AI recommendation. You can develop agency through practice. Biology matters because the uncertainty and discomfort involved can feel harder to tolerate on some days. Understanding that variation helps you decide how to approach a challenge without abandoning it.

What high agency looks like at work

You can be exceptionally good at your job and still let other people decide what your job is. You deliver the requested presentation, meet the deadline and make the changes. Whether the presentation answers a useful question may never come up.

High agency begins when it does.

In George Mack's essay, high agency combines clear thinking, a bias towards action and disagreeability. The last word needs explaining. He means a willingness to resist consensus when it deserves questioning. In a workplace, that might mean asking why a report exists before finding a faster way to produce it.

You recognise these people by what happens around them. Faced with an impossible deadline, they establish which part of the job matters and propose a workable version. When a customer keeps struggling, they investigate instead of forwarding the complaint. They can be junior, softly spoken and thoroughly uninterested in becoming a founder. They still behave as though their judgement belongs in the room.

They also distinguish a genuine constraint from an inherited assumption. 'We tried that once' invites a question about what happened; it doesn't settle whether trying again would be sensible. Their curiosity survives the first objection, and they are prepared to do something with what they discover.

That includes changing their minds. Someone who argues endlessly but never tests an alternative can create plenty of work for everyone else. Agency involves accepting responsibility for what your intervention produces, including discovering that the original plan was better.

AI gives us another authority to question

We've had plenty of practice deferring to authority. School rewards the correct answer, workplaces reward delivering the brief, and social approval makes accepted choices comfortable. Challenging a boss can cost you something. Questioning an AI recommendation carries less social risk, yet its effortless certainty can make disagreement feel surprisingly unreasonable.

The machine has produced six persuasive paragraphs while you're still trying to explain your doubt.

A 2026 experiment with 184 participants puts evidence behind that concern. People solving logic puzzles received advice from a simulated AI system, accompanied by confidence labels. High confidence made participants more likely to accept incorrect recommendations. When confidence accurately reflected reliability, decisions improved (Fregosi et al., 2026). The presentation of certainty mattered alongside the quality of the answer.

At work, the problem goes beyond checking whether a statistic is invented. An AI can produce a beautifully reasoned answer to a question that should never have been asked. A plan to increase meeting attendance may ignore why people find the meetings useless. A polished hiring recommendation may inherit a narrow definition of who would succeed.

High agency means remaining involved in deciding what deserves doing. Before asking AI for a recommendation, write down what you think the problem is and what evidence would change your mind. Then interrogate the answer against those criteria, checking consequential claims at their source. Asking the same model whether it's sure gives it another opportunity to sound sure.

AI can make a tentative idea cheap enough to test. That is an enormous opportunity for someone willing to choose an experiment and take responsibility for the result. Without that willingness, we risk becoming extremely efficient at following recommendations.

Do you have agency, and can you develop it?

Look at your behaviour when the next step hasn't been assigned. Do you investigate, or wait for clarification that nobody has promised to provide? When you disagree, can you propose a way of finding out who's right? You might do this naturally with clients and almost never with your own manager. Agency can look different across the situations you inhabit.

Psychologist Albert Bandura's work offers a useful explanation. Self-efficacy is your belief that you can carry out the actions a particular situation requires. It influences what you attempt and how long you persist. Successful experiences can strengthen that belief; watching someone else manage a challenge can help too (Bandura, 1982).

Think of the first time you question a brief and discover that the client is relieved someone asked. That experience gives you something to draw on next time. 'I can do this' acquires a history.

There is evidence that initiative itself can be taught. A randomised trial involving 1,500 small-business owners in Togo tested training in proactive behaviour, alongside conventional business training and a control group. The initiative programme taught people to anticipate problems and act on opportunities. It increased profits by 30% relative to the control group over two years (Campos et al., 2017). That doesn't promise the same result in your office, but it challenges the idea that initiative is something you either arrive with or lack.

Choose a bounded problem you care about and run a small experiment without waiting to feel transformed. Agree the scope with whoever is affected, then find out what happens. Afterwards, examine how your action changed the outcome. If you manage people, give them decisions they can own and respond constructively when they challenge yours. Asking for initiative while punishing disagreement teaches a very different lesson.

Your comfort zone has a biological context

Stepping outside the familiar involves tolerating uncertainty and the possibility of looking foolish. How manageable that feels can change before anything about the opportunity has changed.

A 2024 meta-analysis covering 154 experiments and 5,717 participants found that sleep loss reduced positive emotion and increased anxiety symptoms (Palmer et al., 2024). This was research on emotion, rather than workplace courage. But it gives us a plausible reason why proposing an untested idea might feel less appealing after a broken night: the emotional conditions in which you're considering it have changed.

Cycle research adds another dimension. A 2026 study of 48 naturally cycling women and 46 men found phase-related variation in motivation during a physical effort task, while measured hormone levels explained little additional variation (Grahlow et al., 2026). It supports paying attention to biological context, without pretending a hormone reading can tell us who will challenge their boss.

Bandura also recognised that people use bodily sensations when judging their capabilities. That creates a trap: feeling more apprehensive today can be mistaken for evidence that you aren't capable of the thing you considered yesterday.

The same reasoning matters when starting a habit or learning something unfamiliar. A growth mindset means believing you can develop; the awkwardness of being a beginner may still feel more tolerable on some days. A rested day with room for mistakes is a sensible time to attempt the demanding first session. On a harder day, rehearse one part or ask someone to work through it with you. This is a planning judgement, not a scientifically established best day to change your life. Keep the commitment while adjusting how far you stretch.

You have agency over your agency

Mack's practical exercise starts with something you value, asks you to identify actions that would express it, then turns one into steps you take. Repeating that process gives you evidence of your ability to affect the world. The idea earns its usefulness through doing something with it.

Our addition is to notice what makes that doing possible. Biology can affect how a challenge feels, and understanding the pattern gives you another choice about how to approach it. You can prepare for a difficult conversation when you have more capacity, or ask for support when you have less, without surrendering the decision itself.

Phase exists to help you build agency, because turning insight into action is about taking back control. Phase ignores your readiness score. It builds a cognitive profile from underlying bio data and maps it against your tasks and calendar to produce an alignment score. Understanding how that differs from readiness helps you use the information to make your own decisions about work.

That is the purpose behind our approach to focus and planning: more understanding of yourself, and more ability to act on it. Join the waitlist and be first in when Phase opens up.

The information in this post is for educational purposes only and does not constitute medical advice. Phase content on health, hormones, and bio data is reviewed by our clinical advisory team. If you have concerns about your health or wellbeing, please consult a qualified healthcare professional.