Everyone wants AI to work. Very few want to fix what it depends on. Tim Hayden made the tension clear. The real blocker to AI value in automotive is not ambition. It is data. For AI to produce meaningful outcomes, the underlying data has to be normalized, labeled, structured, and connected. Not just cleaned. Built to move. Most dealership systems were never designed for that. They were built in layers. Patched together over years. Extended far beyond what they were originally intended to do. The uncomfortable reality is this: if we don't address the structure now, we will pay for it later. Either through costly overhauls or continued dependence on third parties who already have their data house in order. AI is powerful. But power without preparation only amplifies weakness. If the industry wants AI to create leverage, it has to start by rebuilding what sits underneath it.