Am I right that hyperscaling compute is the biggest bet in history? Any counter examples? It’s way more expensive than the Manhattan Project, the Apollo project, and railways across the US. If it doesn’t yield AGI, it may also be the biggest failed bet in history.