Data & Privacy
You Are Not the Customer. You Are the Factory.
Every time you scroll past an ad without clicking it, you have still done exactly what the platform needed you to do.
The Idea
Most people understand, loosely, that free apps make money from advertising. What they miss is the precise mechanism — and once you see it, the entire architecture of the modern internet looks different. The business model is called surveillance capitalism, a term sharpened by Harvard scholar Shoshana Zuboff. The core insight is this: your behaviour is not just a byproduct of using these platforms. It is the raw material being harvested and sold. Every tap, pause, re-read, and ignored notification is logged, processed, and fed into models that predict what you will do next — and those predictions are what advertisers actually buy. This is subtler than 'they sell your data.' Your data is not the product. Your predicted future behaviour is. The distinction matters enormously. Selling raw data is a one-time transaction. Selling behavioural predictions is a perpetual business — because the models get sharper the more you use the service, and sharper models command higher prices from advertisers who want certainty, not guesses. This creates a structural incentive that has nothing to do with your wellbeing. Platforms are optimised to maximise the quantity and quality of behavioural signals you emit — which means maximising your engagement, your emotional volatility (strong feelings produce more signals), and your time on-platform. Your attention is not what they want. Your predictability is.
In the World
In 2014, Facebook ran an experiment on roughly 700,000 users without their knowledge. It manipulated the emotional content of their news feeds — showing some people more positive posts, others more negative ones — to see whether online emotions were contagious. They were. The paper was published in a peer-reviewed journal, which is how the public found out. The outrage that followed focused on the ethics of the experiment. But the more revealing detail was almost overlooked: Facebook had the infrastructure to run this at scale, quietly, as a routine internal test. The emotional manipulation was not the aberration. The capability was the norm. That capability exists because platforms have spent years building extraordinarily granular models of human behaviour. Meta's advertising system, for instance, allows buyers to target people based on inferred psychological traits, life transitions, and purchasing intent — categories derived not from what users declared, but from what their behaviour implied. At one point, internal documents revealed that Meta could identify teenagers feeling 'insecure,' 'worthless,' or 'in need of a confidence boost' — in real time — and flag them as receptive advertising targets. None of this required a villain in a room deciding to harm people. It emerged from a system optimised for a single, coherent goal: turn human experience into data, and data into predictive models, and predictive models into revenue.
Why It Matters
Once you understand this model, 'just don't click on ads' stops making sense as a defence. The system does not need you to click. It needs you to behave — and you cannot stop doing that. What changes, practically, is how you evaluate the services you use. A platform that charges you nothing has made a specific choice about where value flows. That does not automatically make it malicious, but it does mean the design decisions — what gets amplified, what gets buried, what makes you feel urgency or inadequacy — are shaped by incentives that are structurally misaligned with your interests. The more useful question to carry into your digital life is not 'what is this app doing to me?' but 'what does this platform need from me to survive?' When you can answer that, you can start to read the design with clearer eyes — and make more deliberate choices about where you spend your attention, knowing attention is never really free.
A Question to Ponder
If a platform's revenue depends on predicting your behaviour accurately, what does it mean that you keep returning to it — is that a sign it is useful to you, or a sign the predictions are working?
Get a new one of these every morning.
Start learning with Thinkable