Biohacking

Biohacking spans practices whose only constant is that one person is simultaneously investigator, subject and interested party. At one end sit quantified-self measurement, glucose and sleep tracking, and dietary manipulation; at the other, do-it-yourself molecular biology and self-administration of compounds that have never completed a human trial. Absent throughout are the features that make a clinical result interpretable: a prespecified hypothesis, a control condition, blinding and independent measurement.

The culture has a documented boundary case. In 2017 a biotechnologist publicly self-administered a gene-editing construct, and the FDA responded that sale of gene therapy products intended for self-administration is unlawful. The compound repertoire runs ahead of the evidence base, and popularity tracks availability and discussion volume rather than trial quality; several compounds central to peptide biohacking forums have no completed controlled human trial for any outcome discussed there.

Self-experimentation is not worthless in principle. A single-subject design with an objective endpoint, repeated crossover between conditions and a washout can genuinely detect a fast-acting, reversible effect in that person. What it cannot do is detect small effects, slow effects or safety signals, or separate a real change from regression to the mean and the expectancy generated by having chosen and paid for the intervention.

The characteristic misreading is that scale substitutes for design. Thousands of concordant self-reports remain uncontrolled observations sharing the same biases rather than cancelling them, and those who stopped, saw nothing or were harmed are absent from the forum producing them. A second error transplants the software sense of the word: an experiment on a codebase can be reverted, and an endotoxin exposure or an anti-drug antibody response cannot.