If you're deploying agents to do what everyone else is doing faster, you're in a race to zero margin. The question Zero to One forces is whether you're building something genuinely new or just automating the competition.
Peter Thiel's argument in Zero to One is that competition destroys value and monopoly creates it — not the rent-seeking kind, but the kind that comes from building something so distinct that comparison becomes awkward. The move is to own a small, specific market completely before expanding, and to start from a secret: a true thing that most people either haven't noticed or actively resist believing. Secrets are the seed; the monopoly is what grows from them.
The translation is uncomfortable but useful. Agents make it cheap to copy any visible workflow, which means operational differentiation evaporates faster than it used to. Your secret can't be "we automate our pipeline" — that's gone in a quarter. The durable version is a proprietary insight about a specific customer's world that your agent architecture encodes in ways competitors can't reverse-engineer from your outputs. The small-market-first logic gets sharper too: agents let you serve a narrow niche with a depth that was previously uneconomical, which is exactly how you establish the monopoly position before expanding.
- Start from a secret, not a feature — what do you believe about your market that most people would push back on
- own a niche so completely that expansion looks obvious in hindsight rather than optimistic from the start
- treat your agent architecture as the thing that encodes your secret, not just the thing that executes your process.
