A solo episode by one of the hosts, a data scientist. It starts from a family story: his mother and aunt were born deaf from a genetic cause, and he knew he carried the gene. A genetic counselor reasoned only from prior probabilities and ignored the evidence of a hearing first child, a missed Bayesian update.
Biohacking means running N=1 experiments on yourself, guided by your genotype and phenotype. Bryan Johnson is the extreme case; most people stop at supplements, diet, and training. Cheap sensors (CGM, ketone meters) make it easy to monitor practices such as his weekly 40-hour fast.
Whole-genome sequencing ($300) gives you the file to query yourself with LLM-written Python scripts instead of re-testing panels. It revealed statin toxicity, alcohol sensitivity, and fast metabolism of a dental anesthetic.
Caveats: no placebo control, time dependence between trials, and carry-over effects. Still, he argues it points toward personalized medicine, with patients bringing genetic context to fragmented specialists.
Show: explainable
Concepts: biohacking · personal-genomics · bayesian-updating