Here's an interesting way to view the future... I've never seen this argument before, stated in this way.
Have you ever heard about Koomey's law? It makes a strong claim that computations per joule in computing has doubled every 1.57 years since the 40's.
I find it more useful than Moore's law. Why is it a source of deep optimism?
Here's a simple tautology. The only true measure of "advancement" as a species is our growth in computational complexity. So far, that's mostly been achieved through a growth in people via social/cultural advancement. In the last couple of hundred years, we've been doing it through a very rapid increase in our usage of energy. In the last couple of decades, we've begun to use computation to do it.
How does energy increase our computational complexity. Roughly. Typically by eliminating the need for people power to do things in the physical world. Here's a great quote from Hyman Rickover, the father of the nuclear Navy:
"With high energy consumption goes a high standard of living. Thus the enormous fossil energy which we in this country control feeds machines which make each of us master of an army of mechanical slaves. Man's muscle power is rated at 35 watts continuously, or one-twentieth horsepower. Machines therefore furnish every American industrial worker with energy equivalent to that of 244 men, while at least 2,000 men push his automobile along the road, and his family is supplied with 33 faithful household helpers. Each locomotive engineer controls energy equivalent to that of 100,000 men; each jet pilot of 700,000 men. Truly, the humblest American enjoys the services of more slaves than were once owned by the richest nobles, and lives better than most ancient kings. In retrospect, and despite wars, revolutions, and disasters, the hundred years just gone by may well seem like a Golden Age."
Computation replaces both energy and people, in a similar way. We're just starting to use it effectively. The closer we align our economy to take advantage of computation efficiency improvements, the fast we progress.
The ongoing growth in computational efficiency unmoors us from a reliance on energy growth. It does so in much the same way energy growth has disconnected us from population. In fact, given the amazingly fast rate of improvement in computational efficiency we are seeing, we could see substantial reductions in both population and energy production in the future w/o seeing any slowing in the growth of the computational complexity of our species.
Of course, that's not possible in the current substrate. We're going to need to see a break with the past as profound as we saw with our break agrarianism. Eventually, we'll see a system that aligns with the potential for improvement that is possible and just as we saw in the contest between industrial nations and undeveloped nations in the last three centuries, it won't be a fair fight. The only lingering question is how rocky the transition is and what the new vehicle looks like.
One way to accomplish this quickly (before the old system completely fails) is to decentralize production with networked resilient communities. At that point, the global economy would be 98% information and all basic needs would be provided for locally. Information only economies built on a base of self-sufficiency can be as different from capitalism as capitalism is from feudalism.