Agency, AI & the One-Person Billion-Dollar Company
I was doom-scrolling this weekend when Gian Segato’s “Agency Is Eating the World” flashed by. Ten minutes later half the essay was highlighted, and my coffee was ice-cold.
Segato argues that agency – the raw determination to act without waiting for permission – has become the decisive fault-line in an AI world. I couldn’t agree more. In “The Solo Billionaire” I’d asked whether the first one-person unicorn was already hiding in plain sight, and the evidence has only strengthened since.
Specialization -> Generalist Leverage
For decades deep domain expertise was a local monopoly: spend ten years perfecting an esoteric skill, defend the castle, and bill by the quarter hour. Today a $20 ChatGPT subscription drafts the patent, builds the go-to-market model, and designs the CRISPR guide in minutes. Expertise still matters, but it’s no longer the bottleneck; the scarce resource is the leader who can see across silos and orchestrate tools that were unimaginable a year ago.
In practical terms, losing that monopoly blows the talent market wide open. When anyone can summon a first-draft patent or working financial model before lunch, the real premium shifts to people who can stitch those outputs into a coherent plan. Idea velocity spikes because prototyping now costs coffee-money; a weekend hack can replace months of cross-department hand-offs. Capital efficiency follows – tiny seed rounds can fund a minimum-viable company, not just an MVP. Careers flatten into portfolios of micro-skills, and teams assemble fluidly across geographies because the heavy lifting is done by shared AI tooling. The gate that once rationed innovation – “Do you have the credential and the headcount?” – swings wide; the new question is “Can you frame the problem and orchestrate the tools before the next person does?”
Solo Scale Is No Longer Theory
I wrote last year that solo founders would soon run nine-figure businesses. Latest proof-point:
“Cursor writes almost 1 billion lines of accepted code every day.” – Aman Sanger, co-founder of the four-person team behind Cursor’s AI editor.
“Accepted” doesn’t mean flawless – just that a human hit Cmd + Y and shipped it – but the number still vaporizes any comfy notion of linear productivity. When four engineers can out-code a mid-sized country, the constraint flips from typing skill to imagination and prioritization.
The Personal Operating Layer Gets Teeth
In a previous post on the Personal Operating Layer (POL), I described AI that remembers, verifies, and acts so humans can operate at intention-speed. Agency is the behavioral twin of the POL: once your digital chief-of-staff can book clinical-trial patients or simulate your Series A cap-table, the only question left is what do you want to build? Tools no longer constrain imagination; imagination constrains outcomes.
Implications – Healthcare, Life Sciences & Beyond
Boards & C-Suites: Hire integrators who speak biology and Python. They’ll unlock 10-X leverage on existing assets.
Operators: Stop polishing job descriptions and start shipping prototypes. Your advantage is the speed at which you test hypotheses, not the headcount you supervise.
Investors: Office-space and FTE headcounts are obsolete heuristics. Look instead for proof-of-agency: founders who bend models to their will and post real traction on a burn rate that rounds to zero. I’ve seen plenty of ‘smart,’ ‘cute,’ even downright brilliant ideas – none of it matters if you don’t nail product-market fit. Get that right and brilliance takes care of itself.
Policy Makers: Credentialism won’t disappear, but it will be challenged. The regulatory moat moves from “Who has the degree?” to “Who can demonstrate transparent, auditable AI pipelines that manage risk?”
Segato says his worldview has collapsed into a single bit: agency or no agency. I’d phrase it this way: in a world where tools are universally abundant, the scarcest asset is conviction married to execution. If the cost of action has dropped to near-zero, what’s the boldest move you’ll make next?