You Are in Control of Your State: Why Human Outcomes Are Controllable Through Causal State Intervention

A new theoretical framework argues that moment-to-moment human outcomes are controllable through interventions targeting a person’s dynamic latent state, challenging the long-standing puzzle of within-person variability in behavioural science [1]. The framework, posted to arXiv on 26 May 2026, defines a person’s “state” as a time-indexed weighting vector over biological, physiological, and neuropsychological dimensions that govern how the next event is processed into a decision and an outcome [1]. The authors assert that the relationship between state, decision, and outcome is causal, not merely correlational, and that this weighting vector shifts at sub-daily timescales [1]. The conscious channel through which outcomes become reportable is described as a narrow attentional bottleneck whose contents are themselves state-dependent [1]. The work draws on six established strands of evidence: causal inference, predictive processing, allostasis, the attentional bottleneck, chronobiology, and computational psychiatry [1]. It is also grounded in 24 months of observational data collected from a deployed behavioural platform that included more than 200,000 consented users across four occupational personas, spanning a research period from 2023 to 2026 [1]. The central claim is that the same individual, given the same observable input, can produce different outcomes on different occasions because the latent state differs [1]. The authors propose that this variability is not noise but a target for precise intervention. By adjusting the state trajectory at the moment a decision is being formed, they argue, an outcome becomes conditionally controllable [1]. The paper derives seven testable predictions and lists six operational requirements for building what it calls state-aware systems [1]. It discusses potential applications in digital health, education, AI personalisation, and personal agency [1]. The preprint has been indexed with links to code and data tools including alphaXiv, CatalyzeX, DagsHub, and Hugging Face [1]. While the framework does not engage with philosophical questions about free will or determinism, the broader context of human agency has been debated across disciplines for centuries. The question of whether individuals can steer their own outcomes touches on long-running inquiries into the meaning of life and the nature of personal control, topics that have generated extensive philosophical, scientific, and theological speculation without consensus [5].