Uploaded July 2026 | Updated September 2026, 25 minutes ago
Why your AI writes 70 pages when you just needed four.
Chris McClanahan, a product lead over Tanzu Platform, on why AI output goes overboard and how to rein it in. He fed an agent four paragraphs of intent and got 70 pages back. Thorough, and useless. “If I were to send that to anybody, they would laugh at me.”
The fix is boundary conditions: tell it four pages instead of seventy, and for the UI work, hand it the actual CSS file and the list of components it’s allowed to use. Constrain it and the output snaps back to something you could ship. The more you refine the edges, the better the product on the other side.
From a longer conversation with Coté about how Tanzu’s product team rebuilt itself around AI agents.
Interested in a great platform as a service for your enterprise apps? Try Tanzu: vmware.com/products/app-platform/tanzu
Why your AI writes 70 pages when you just needed four.
Chris McClanahan, a product lead over Tanzu Platform, on why AI output goes overboard and how to rein it in. He fed an agent four paragraphs of intent and got 70 pages back. Thorough, and useless. “If I were to send that to anybody, they would laugh at me.”
The fix is boundary conditions: tell it four pages instead of seventy, and for the UI work, hand it the actual CSS file and the list of components it’s allowed to use. Constrain it and the output snaps back to something you could ship. The more you refine the edges, the better the product on the other side.
From a longer conversation with Coté about how Tanzu’s product team rebuilt itself around AI agents.
Interested in a great platform as a service for your enterprise apps? Try Tanzu: vmware.com/products/app-platform/tanzu
