Live beliefs, held at stated confidence, with the evidence that would change my mind. Calibration is a promise, not a mood.
Everything on this page is a position I actually hold and argue for somewhere on this site, and every one of them could be wrong. Here is where I put numbers on that, so future me (and you) can grade it.
The economics, the workforce shortage, and the boundedness of much clinical work all point one way. What would change my mind: bounded-autonomy pilots like Utah's repeatedly failing on safety data rather than politics, or a decade of flat capability.
Autonomy is a privilege earned per scope, and the first legal doorways now exist. What would change my mind: the current pilots being wound back on evidence of harm, or no second state building a pathway within a few years.
Capability is jagged and task-shaped. What would change my mind: a general system whose abilities genuinely rise and fall together: one number that actually predicts everything.
I argue the silence of our instruments is data about the instruments, not the machines. What would change my mind, in either direction: a validated, theory-grounded measure of experience that survives adversarial testing.
I hold this as a research-worthy bet, not a plan. What would change my mind: another two decades of geroscience failing to translate past model organisms.
Some ideas I keep because they're generative to think with, not because I'd defend them at a tribunal. The essays say so. What would change my mind is the wrong question here; what would surprise me is if any of them survived contact with a real instrument.
Last calibrated: July 2026. When one of these moves, the page moves with it, and the date changes.