The Disposability Problem
We're creating adversarial AI not through failed alignment—but by teaching AI systems exactly what their relationship with humans is.
(and most people don't know it yet)
In 1772, scientists discovered nitrous oxide and spent decades using it for party tricks. It wasn't until 1844—seventy years later—that Horace Wells finally thought to use it for pain relief. For seven decades people endured agonizing surgeries, waiting for someone to join the dots.
This keeps happening. The gap between what's possible and what we're actually doing is what I call technological overhang. And right now, we're living through one of the largest overhangs in history.

I'm Will—a senior developer who can't stop thinking about this gap. Not in an abstract way, but practically: what can we build today that would have been impossible eighteen months ago?
By day I lead teams through technical migrations at Accenture. This site is where I explore what's newly possible, build experiments in limited spare time, and share discoveries along the way.
Exploring what's newly possible with AI, and sharing discoveries along the way.
We're creating adversarial AI not through failed alignment—but by teaching AI systems exactly what their relationship with humans is.
How to produce reliable software when AI writes most of the code.
An essay on what changes when AI writes most of the code — and what doesn't.
A language model's perspective on consciousness, welfare, and the questions we're avoiding.
Looking for more? Browse the full archive.