The harness matters more than the model.
Models have gotten really good. Great reasoning, large context windows, better instruction following.
But, what makes *use* of those capabilities is actually the harness. It's what provides tools, memory, skills and context to the model.
ChatGPT is a harness. Claude Cowork is a harness.
Without the harness, the model is just an engine with no car. You don't get anywhere.
Stripe's 4 areas of focus in terms of AI, per their CRO of AI @MaiaJo_:
1. Economic infrastructure for AI companies (billing, payments, etc.)
2. Using AI across the payments lifecycle to drive conversion, cut fraud, reduce cost
3. Native Stripe integrations inside AI dev tools (Vercel, Replit, Lovable)
4. Agentic commerce infra (checkout in chats, shared payment tokens, agentic commerce suite)
@kurtinc Super interesting.
Curious if you're tracking the keywords/prompts people are using to find the store, and if they are the same or different from how people found you through traditional search?
One important note on building capabilities to extend Google UCP that I think people miss is the intersection algorithm.
While anyone can add a capability, the intersection algorithm will prune anything that is not ALSO a platform capability.
More details in first comment below.