This article is part of "Minds in the Machine Age" — a companion series to Between Brain & Binary.
← Back to Series Overview | ← Previous: Creativity Isn't Computation: What AI Art Teaches Us About Ourselves | → Next: Trust in the Black Box: Why Explainable AI Matters
You have tried to build the same habit fourteen times. You know this because the apps on your phone know this - the meditation streaks that run to eleven days before breaking, the fitness challenges completed through week two and abandoned in week three, the reading goals set in January that quietly lapse by February. Each attempt begins with genuine intention. Each lapse comes with genuine disappointment. And somewhere in the middle of all of it, surrounded by more behaviour-change technology than any generation in history, you wonder whether you are simply not the kind of person who sticks with things.
You are. The tools, in many cases, are not set up correctly.
What Habits Actually Are
A habit is not, in the first instance, a decision. It is a learned behaviour that has become automatic through repetition - executed with little conscious deliberation, triggered by cues in the environment rather than by deliberate intention. Charles Duhigg, drawing on MIT research into the neuroscience of habit formation, described the underlying structure as a habit loop: a cue that triggers a routine that produces a reward, with the reward reinforcing the cue-routine association until the sequence becomes self-executing.
This architecture has a significant implication: habits are not primarily about motivation. They are about context. The most powerful variable in whether a behaviour becomes habitual is not how much you want to do it - it is how reliably the cue appears and how frictionless the routine is to execute.
BJ Fogg's Fogg Behavior Model formalises this insight. Behaviour happens, he argues, when three elements converge at the same moment: motivation, ability, and a prompt. Change any one of those variables and you change the likelihood of the behaviour. Fogg's central practical finding - developed through his Tiny Habits research at Stanford - is that motivation is the least reliable lever. The more dependable path to lasting behaviour change is increasing ability (making the behaviour easier) and engineering better prompts (embedding cues into existing routines).
A 2010 study by Phillippa Lally and colleagues, published in the European Journal of Social Psychology, put a more precise number on habit formation than the commonly cited (and incorrect) "21 days." Across the 96 participants studied, the median time for a new behaviour to become automatic was 66 days — with significant variation depending on the behaviour complexity and individual differences. Some habits formed in 18 days. Others took 254.
The 21-day myth, perpetuated by decades of self-help culture, is not just inaccurate. It is harmful - because people who do not see automaticity by day 22 conclude that they have failed, rather than that they are simply still in the formation period.
What Technology Gets Right and Gets Wrong
Behaviour-change apps have become sophisticated at certain things. Streak counters leverage loss aversion — the asymmetry, first formalised by Daniel Kahneman and Amos Tversky in their 1979 prospect theory paper, in how much more we dislike losing something than we enjoy gaining an equivalent thing — to make showing up feel low-stakes but missing feel consequential. Notifications serve as external prompts when internal cues have not yet formed. Progress visualisations satisfy the psychological need to see evidence of movement.
But most habit apps are also built on a flawed model of motivation. They assume that engagement - opening the app, completing the check-in, watching the streak grow - is the goal. Nir Eyal's Hooked framework, which underlies much consumer app design, optimises for return visits and variable reward schedules: the same mechanisms that make social media compelling. These are not the same mechanisms that produce the kind of intrinsic motivation that sustains long-term behaviour change.
Research on self-determination theory, developed by Edward Deci and Richard Ryan, consistently shows that behaviours driven by intrinsic motivation - doing something because it is genuinely meaningful, enjoyable, or aligned with personal values - are dramatically more durable than behaviours driven by external rewards or social pressure. An app that makes exercise feel like a point-collecting game may increase short-term engagement while undermining the development of the intrinsic connection to exercise that would sustain it without the app.
The habit technology that works best tends to be the technology that eventually makes itself unnecessary.
How AI Could Actually Help
The genuinely promising applications of AI to behaviour change are less glamorous than the streak counter, but more aligned with the actual science.
Peter Gollwitzer's research on implementation intentions - the simple "when X happens, I will do Y" formulation - shows that specificity of planning dramatically increases follow-through. People who decide not just to exercise but to exercise at 7am on Tuesdays and Thursdays in the park near their house are significantly more likely to actually exercise. AI tools that help users build these specific, contextualised plans - and that adapt them as life changes - are working with the grain of behaviour science rather than against it.
One design pattern that does align with the evidence is habit stacking - anchoring a new behaviour to an existing, already-automatic one, so the established habit itself becomes the cue rather than relying on willpower or an external notification to supply it. The technique has no single origin point - it draws on the same cue-routine-reward architecture Duhigg described - but James Clear's popularisation of it captures why it works better than most app-based prompts: an existing habit is a more reliable trigger than any notification, because it is already wired in and does not compete for attention against a phone full of other notifications making the same bid. An AI tool that helps a person identify which existing routines could anchor a new one is doing something structurally different from a tool that simply reminds them at 7am and hopes.
AI can also offer something that static apps cannot: genuine responsiveness to context. A system that notices you have missed three Monday sessions and asks what changed - and adjusts the habit structure accordingly - is doing something closer to what a skilled coach would do. A system that identifies that your habit completion rate drops sharply on travel weeks and suggests a travel-specific alternative is working with your life rather than against it.
What most current AI habit tools have not yet cracked is the underlying motivation architecture. They can make the habit easier. They can improve the prompt. What they cannot yet reliably do is help you understand why this matters to you - in the deep, identity-level way that predicts long-term change.
The Self You Are Building
James Clear argues that the most effective framing for habit formation is identity: not "I want to exercise" but "I am a person who exercises." The habit is the evidence. The repetition is the vote you cast, daily, for the version of yourself you are choosing to become.
No app does that for you. The technology can support the structure. The meaning is something you bring.
The question worth asking, before downloading the next habit tracker, is not whether this app will make the behaviour easier. It is whether you have yet got clear about why the behaviour matters - and whether the tool in your hand is designed to deepen that clarity, or simply to keep you swiping.
References (APA style)
Clear, J. (2018). Atomic habits: An easy and proven way to build good habits and break bad ones. Avery.
Deci, E. L., & Ryan, R. M. (2000). The "what" and "why" of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268.
Duhigg, C. (2012). The power of habit: Why we do what we do in life and business. Random House.
Eyal, N. (2014). Hooked: How to build habit-forming products. Portfolio/Penguin.
Fogg, B. J. (2009). A behavior model for persuasive design. Proceedings of the 4th International Conference on Persuasive Technology, 1–7.
Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493–503.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291.
Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009.
AI Disclosure: Research and organization for this article were assisted by AI tools; all factual claims and citations were independently verified against primary academic sources, and the analysis and conclusions are the author's own. The featured image was generated using Canva AI.
← Previous: Creativity Isn't Computation: What AI Art Teaches Us About Ourselves | → Next: Trust in the Black Box: Why Explainable AI Matters
Related reading from Between Brain & Binary: Looking Ahead: The Future of Mind and Machine
Comments
Post a Comment