When people talk about “the algorithm”, they often mean it as a single, shadowy entity – something opaque, powerful, and vaguely intentional.
That framing is unhelpful.
At Luminae, when we refer to “the algorithm”, we are not talking about a conscious system, a moral agent, or a hidden intelligence making decisions about you. We are talking about a set of automated processes designed to do one thing extremely well: predict what will keep you engaged.
Understanding that distinction is the starting point for retaining agency.
What “the Algorithm” Actually Is
In practical terms, an algorithm is a collection of rules and models that process signals and make predictions.
Those signals are generated by your behaviour:
– what you watch or read
– how long you linger
– what you skip, save, search for, or share
– how you react, and how quickly
From this, platforms build a statistical profile. Not a picture of who you are, but a working approximation of what you are likely to do next.
The system does not understand your values, intentions, or context. It understands patterns. If certain content reliably holds your attention, it is shown more often. If it does not, it fades away.
Over time, this produces a feed that feels personal, even though it is driven by probability rather than insight.
How the System Learns Without Asking
One of the most important features of modern recommendation systems is that they do not need explicit input.
You are rarely asked what you want your feed to become, what kind of thinking you wish to cultivate, or what long-term outcomes you value. Instead, the system infers preference from behaviour.
This means learning happens continuously and silently.
Small actions matter:
– a pause on a video
– a half-read article
– a late-night scroll
– a moment of curiosity
None of these choices feel consequential in isolation. In aggregate, they are highly informative.
From them, the system learns not just topics, but tendencies: emotional tone, complexity tolerance, novelty appetite, and sensitivity to reward or reassurance.
Why “It Knows You” Is a Misleading Phrase
Saying the algorithm “knows you” gives it too much credit.
What it knows is a narrow slice of you – the part that shows up through measurable interaction. Reflection, restraint, doubt, and private reasoning leave little trace.
The risk is not that the system understands you deeply. The risk is that its partial understanding becomes the environment you think within.
When certain ideas, narratives, or identities are repeatedly surfaced, they begin to feel more salient. When others rarely appear, they quietly recede. This shapes what feels available, reasonable, or worth considering.
No persuasion is required. Familiarity does the work.
Why This Matters More Than We Assume
The effects of algorithmic curation are rarely immediate or dramatic.
They show up over time as:
– shifted baselines of what feels normal
– increased confidence in unexamined assumptions
– narrowing of perceived options
– a sense that certain paths are simply “what people like me do”
Because these changes occur gradually, they are easy to mistake for personal preference or independent choice.
This is where agency quietly leaks away – not through coercion, but through unexamined reinforcement.
From Prediction to Direction
A system built to predict engagement inevitably begins to influence behaviour.
When content that holds attention is consistently prioritised, the environment starts to reward certain reactions, interests, and identities more than others. Over time, this nudges behaviour in predictable directions.
This does not require intent. It is an emergent property of optimisation.
The crucial shift is recognising when prediction turns into direction – when a system designed to respond to behaviour starts shaping it.
What Awareness Changes
Once you understand how “the algorithm” actually functions, several things become possible.
You can begin to:
– notice patterns rather than individual posts
– recognise when a theme or desire is being reinforced
– question why certain content feels increasingly prominent
– separate genuine interest from repeated exposure
This awareness does not demand rejection of platforms or withdrawal from digital life. It simply restores a pause between stimulus and response.
That pause is where choice lives.
The Luminae Position
Luminae’s position is neither alarmist nor complacent.
Algorithms are powerful tools. They can surface useful information, accelerate learning, and support chosen goals. But they are indifferent to your long-term wellbeing, values, or development.
They optimise for engagement, not understanding.
The responsibility for direction therefore sits with the individual – provided they are given the insight needed to see how direction is being applied.
That is the purpose of Luminae’s work: not to warn people away from influence, but to help them recognise it clearly enough to decide.
In the next Insight Library post, we will move from theory to practice, exploring the subtle signals that indicate when algorithmic influence is shaping desire – and how to reintroduce friction before momentum takes over.
Explore the connections
Explore the research
Academic research notes supporting this Insight are available on request. Use our contact form and tell us which topic interests you.