Engineering the Future

For a long time, the people I loved mattered to me. My children mattered. The life I was damaging mattered. The problem was not an absence of love or information. The problem was that, in crucial moments, those things could stop competing effectively with what was immediately available. Tomorrow still existed, but it didn’t feel reachable enough to govern the next hour.

I’m a father of two teenagers and an engineer trained to think about turbulence, instability and nonlinear systems. I also spent a year and a half in a therapeutic community. Those identities do not make my experience universal, and my biography is not scientific evidence. They do, however, shape the question I keep returning to: what changes when a person begins to recover, and what begins to change before a return to use?

In the community, recovery did not arrive as a single decisive insight. A predictable day, other people, repeated responsibilities and fewer hidden spaces carried parts of life that my conviction could not yet carry alone. Agency grew inside structure. Gradually, actions that had felt remote became possible again – keeping an appointment, telling the truth earlier, tolerating discomfort, imagining a future and doing something small on its behalf.

Relapse is often narrated as a switch – someone was doing well, then made a bad choice. Choice and responsibility matter, and a ‘systems view’ must never be used to excuse harm. But the switch story is incomplete. The visible event may be the end of a trajectory in which sleep has deteriorated, contact has narrowed, shame has become harder to disclose, cues have intensified and ordinary routines have lost their grip.

In engineering, separate pressures can interact so that the response is far larger than any one input would suggest. A system can even look stable in the long run while showing strong temporary amplification. I use that idea only as an analogy. It’s not a demonstrated brain mechanism, not a clinical model and not a licence to translate equations from fluid dynamics into a person. Its value is narrower: it asks us to look for interaction, timing and context rather than a single cause.

Poor sleep may not be decisive on its own. Nor may loneliness, money stress, conflict, cue exposure or an unstructured weekend. Together, in a depleted recovery environment, they may produce a much larger response. The same trigger can be small in a robust field and explosive in a depleted one. That possibility changes the practical question from ‘Which risk factor caused this?’ to ‘What combination was gathering, and when did the route back begin to narrow?’

Recovery is often treated as something stored inside an individual – motivation, insight, discipline or readiness. Those things matter, but they are not the whole field. Housing, income, relationships, belonging, access to care, meaningful activity and the ability to return after a difficult day all change what actions are actually available. This is close to the idea of recovery capital, but I find it helpful to make the idea dynamic. Resources do not simply sit in a list. They reinforce or weaken one another over time.

A supportive conversation may restore enough trust for someone to attend an appointment. The appointment may restore enough stability to sleep. Better sleep may make it easier to resist a cue or disclose a lapse before it expands. The direction can also reverse. Isolation makes shame easier to hide; shame makes contact harder; lost contact leaves fewer interruptions between impulse and action. What looks like one person failing may also be a field losing its capacity to return.

The attraction of systems language is also its danger. A neat score can create false confidence. I have proposed a ‘recovery coupling index’ as a research scaffold for thinking about interacting pressures, but it is not a validated relapse predictor and should not be presented as one. People are not machines, and early-warning signals are probabilistic, incomplete and deeply dependent on context.

Still, we can notice changes without claiming certainty. Has the person stopped answering the one contact they usually trust? Has sleep shifted for several nights? Are ordinary commitments becoming harder to start? Is the future horizon shrinking from weeks to hours? Is a lapse being hidden because re-entry feels humiliating? None of these observations diagnoses an outcome. But together they may tell us that the window for a low-friction return is becoming smaller.

For families, peers and services, this suggests a different kind of conversation. Instead of waiting for proof of collapse, we can ask what has recently become less reachable. Which relationship still has access? What’s the smallest honest contact that can happen today? Which practical pressure can be reduced? What would make returning easier without removing accountability?

To say that behaviour emerges from interacting conditions is not to say that nobody is responsible. Harm remains harm. Families may need boundaries, and safety may require distance or professional help. A compassionate systems view does not ask other people to absorb unlimited damage – it asks whether we can hold responsibility while intervening earlier and designing fewer humiliating barriers to return.

Flow Hijacked

Tomer Avraham is the creator of Flow Hijacked, a free, independent and noncommercial bilingual project that connects lived experience with carefully bounded scientific models. The material is educational and is not a substitute for medical care, diagnosis or crisis support. flowhijacked.com/go/ddn