Having lots of money to throw at moonshot problems is always important. The current breakthrough of the 2020s is LLMs. I wonder if such breakthroughs can be achieved with this kind of approach.