I agree it's directionally correct, but only in the ways that don't matter to this discussion. If 2026->2029 AI is as much of an improvement as 2023->2026 AI, is anything we learn about how to leverage it in 2026 going to stay relevant?