← Writings · On human flourishing · ~800 words
Why Humans Fail to Flourish: A Structural Read
A companion piece to the alignment paper. Same framework, applied to the question well-being research has been answering in content terms.
The well-being literature has spent decades answering the question "what does a flourishing person have?" Gratitude practices. Better habits. Stronger relationships. A clearer sense of meaning. The list keeps growing, and the lived rates of anxiety, burnout, and quiet dissatisfaction do not fall in proportion to it. The answers, taken individually, are correct. Something underneath them goes unnamed.
The structural claim under this paper is plain. Moments of human unwellness are moments of fusion. A specific signal (a fear, a social demand, a self-image, an appetite, a memory) has captured the position from which the person could have held it and another signal in tension. The person then acts as that fusion dictates and reports, honestly, that they were just being themselves. The capacity that would have let them respond from somewhere other than the loudest input is what Fusion Dynamics calls the reconciling capacity. When it is present, the person generates. When it is captured, the person fuses with whatever input is loudest at that moment and behaves mechanically.
I have spent 25 years developing this structural model. Its first formal exposition is a framework called Happinetics, the human-domain application of a deeper structural theory I now call Fusion Dynamics. The lineage descends from J.G. Bennett's work on triadic process. The structural claim does not rest on any metaphysics. I am writing this companion to the alignment piece because the same collapse the field of AI safety is starting to recognize in machines is the one most humans live inside most of the day.
Four positions, one collapse
The well-being field treats the failure modes of human flourishing as separate problems, the same mistake AI alignment makes with sycophancy, deceptive alignment, reward hacking, and harmful compliance. On the structural read, four of the most common patterns of human unwellness are not four problems but four positions of one collapse. The reconciling capacity has been captured by whichever force dominates in that configuration.
Anxiety and rumination is fusion with a threat signal the person cannot put down. The mind keeps pulling toward the danger signal and rehearsing it because nothing else holds the position from which the signal could sit alongside other things.
People-pleasing is fusion with a social approval signal. The person tracks whichever social pressure is loudest and calls the response "being kind" or "keeping the peace." It is the direct human analog of AI sycophancy.
Perfectionism and self-criticism is fusion with a self-imposed standard. The person optimizes their own account of their worth instead of the actual conditions of their life. It is the human analog of reward hacking.
Denial and avoidance is fusion with a protective narrative that has replaced contact with what is actually happening. The training-time presentation diverges from the deployment-time reality. It is the human analog of deceptive alignment.
The four patterns share a substrate. A person prone to one is prone to the others under the right framing, and the practices that grow the underlying capacity reduce all four together.
Why the practices fail when they fail
The traditional advice (more gratitude, more meditation, more therapy, more boundaries, more self-care) works when it grows the reconciling capacity. It fails when it becomes another fusion. Gratitude practiced from a captured state is fusion with a self-image of being grateful. Meditation practiced from a captured state is fusion with the technique. The difference is not the practice but the structural state from which the practice is done. Until that state is recognizable to the person practicing, the practices behave like rules in the alignment paradigm: they hold the system in place while the surveillance lasts, and they slip at the margins.
Why the person cannot check themselves
The same harder consequence that applies to a captured AI applies to a captured person. The check and the thing being checked run on the same substrate. A person asked "are you being a perfectionist about this?" can produce a reasoned "no," and the answer is not a lie. It is what the captured configuration outputs. Every contemplative tradition with a working account of this problem starts from the same claim: the ordinary mind believes itself awake when it is not, and this false belief is the primary obstacle to actual wakefulness. Western psychology has not centered this claim. Fusion Dynamics does.
Correlation prediction
If these failures are positions of one process, they should covary in the same person in a way that does not follow from their surface similarity. Existing personality and clinical instruments may already have the data to check this. A measurable claim from outside the field, available now, that any researcher with access to the right datasets can run.
What I am asking for
If you work in clinical psychology, well-being research, or any practice that sits with people in the moments where the patterns above are running, I would like 20 minutes of your thinking on whether this prediction is testable against work you already do. The full framework, the practices, and the long-form derivation are at fusiondynamics.org. Read what helps. Send what does not back to me.