AI Content, Dependency & Digital Wellbeing

When does engagement become something else?

AI is becoming increasingly good at learning what holds our attention.

Recommendation systems learn from what people watch, click and return to. Generative AI takes this further by allowing content itself to be created, personalised and adapted in response to those signals.

The result can be a continuous feedback loop between content, behaviour and optimisation.

But there is something traditional engagement data struggles to tell us:

What is happening to the relationship?

Thetafan is developing patent-pending AI to measure relationships between people and content, including how signals associated with fandom, trust, loyalty, influence and dependency change over time.

Engagement isn't the same as dependency

Someone returning frequently to content doesn't necessarily have an unhealthy relationship with it.

Equally, measuring watch time, clicks or sessions alone may not explain how a person's relationship with that content is changing.

This distinction matters because AI systems are increasingly capable of adapting experiences around individual behaviour.

A system can know that engagement increased.

Understanding why it increased, and what kind of relationship developed alongside it, is a different measurement problem.

That is the problem Thetafan is working on.

Measuring the relationship

Thetafan began with fandom.

The original challenge was to understand how multiple behavioural signals could distinguish a casual audience member from someone developing a deeper relationship with content, a creator or a community.

The same underlying technology creates the potential to understand a broader spectrum of relationships.

Fandom. Trust. Loyalty. Influence. Dependency.

Rather than treating these as individual interactions, Thetafan analyses signals over time to build an evolving picture of the relationship behind the behaviour.

Why this matters for children

Children and young people deserve particular attention as AI-driven experiences become increasingly personalised.

The question isn't simply how much time a child spends online. Context matters, and high engagement does not by itself demonstrate harm.

The more useful question may be how that person's relationship with the content is changing over time.

Thetafan is researching how relationship intelligence could provide platforms and governance stakeholders with another measurable signal for understanding those changes, identifying patterns that warrant attention and assessing what happens after an intervention.

Digital wellbeing needs better measurement

Digital wellbeing cannot be understood from a single metric.

Time spent, engagement and retention provide useful information, but they were not designed to describe the full relationship between a person and the content they experience.

Relationship intelligence adds another perspective.

For content owners, it can help distinguish an audience from a fandom.

For communities, it can help understand trust, loyalty and influence.

For platforms and governance teams, it could help identify when relationships are developing in ways that deserve closer examination.

The underlying technology is the same.

The application depends on the relationship being measured.

Building a measurable layer for the AI content era

Thetafan is developing technology for a world in which content may be created by people, selected by algorithms and generated by AI.

As those systems become better at understanding human behaviour, we believe organisations will need better ways to measure their effects.

Not simply:

“Did someone engage?”

But:

“What kind of relationship did that engagement create?”

Explore Thetafan

Thetafan's proprietary, patent-pending technology measures relationships between people, content and communities, from commercial fandom and loyalty to emerging applications in digital wellbeing and platform governance.

If AI can optimise the content, we should be able to measure the relationship it creates.