Research
My day job is automating work, and it keeps raising a bigger question: what do AI and automation actually do to labor and to efficiency? This is where I study it, from the economics side and from the AI side.
The questions I keep coming back to
- How much time does automation really free up, and where does that time go?
- Which tasks can AI handle reliably, and which still need a person?
- When work changes, who gains and who absorbs the cost?
Economics research
Why economics: automation is an economic change before it is a technical one. I want to know who gains and who pays when the way work gets done changes, through labor, prices and productivity.
Nothing published yet.
Go to economics researchAI research
Why AI: I think of a model as a coworker that has to stay reliable across long stretches of work, not just single answers. This is where I test what today’s models can do on real workflow tasks, and when a person should stay in the loop.
Nothing published yet.
Exploring next · IdeaLong strategy board games, like Axis & Allies, as LLM benchmarks: can a model hold a plan across hundreds of turns?
Go to AI researchEverything, newest first
Nothing published yet. Each entry will state its question, method, data and status up front.