You may be right, especially with the growth of applications of software. Personally I'd rather bet on slower revenue growth than the current 30%. Not necessarily much slower, but even 25% yearly growth over 10 years would be a big difference in the end compared to 30%. My thinking is that usage of cloud compute can grow greatly, but with revenues growth lagging behind, because of falling costs of compute (more powerful/efficient CPUs etc), economies of scale, and competition putting pressure on prices. For example AWS operating margin is 34% currently, I expect that to fall as the market matures (but Google's cloud margins are much lower right now).