Open source, and execution as the moat

The argument

• The technology was never the moat. Dinis: open source everything "because the technology was never the moat"; what is defensible is "usually the data, the relationships, the distribution, or the speed at which you ship" (The question is whether they miss it, 21 September 2026).
• The moat is a rate, not a wall. "A competitor who forks our code today gets our position as of today. They do not get our velocity." And: "The code is open source. The execution is not forkable." (open-source.sgit.ai).
• What is sold is trust. "We are selling trust" (same site). A buyer, a regulator or an auditor can read every line. For agents that act on someone's behalf, that is the product.
• Customisation is the business. Open code is the evaluation version; what is sold is making it work for one buyer (The value of constraints, 16 October 2026; Selling less code, 30 October 2026).

What it protects for an investor

Dinis's view, in The fork next door (16 October 2026): because anyone can fork, founders, customers and investors are each protected from the others. A customer is never locked in; a founder who stops executing can be replaced by the fork; an investor's protection is the work in the open, not a clause. In Sell, don't issue (16 October 2026) he goes further: open source as the investor's protection instead of control rights, and he lists that idea's costs himself.

The honest weakness

The open-source site itself calls "execution is not forkable" its weakest, most contestable claim. A well-funded competitor could copy the code and hire faster. Our answer is to keep the pace visible, so it can be judged: every release, every commit, every decision is public (versions). If the pace stops, the argument stops with it. That is why we report what we do, not what we expect.