Hi; a distribution of probabilities, indeed.

• At one extreme, fusion-powered nanotech hyper-AI technological event horizon (even if "the future is here, it's just not evenly distributed" - w. gibson )
• In the red corner, orlovian stage 5, mad-max, sudden mass crash, cannibalism, liebig minima.

Bewtixt the two lie :

• return to permanent growth
• BAU, boom-bust business as usual
• very slow decline, permanent mild recession
• .. and the orlovian stages 1..4

But what are the odds?

We are clearly in Orlov-stage-1 now, and risk stage-2 if growth returns and energy prices spike up again.
There's no sign of a return to long-term growth, and people like Roubini suggest the odds favour a long-term depression.
I'd give the techno-nirvana scenario slim odds, though keeping an eye on what happens with Brussard's polywell.
The ecological models of overshoot and collapse seem strongly predictive, otherwise, and support the gloomier outlooks.

I think the systems are so complex with so many types of feedback and gaming going on that any certainty would be an illusion.

But at the same time there's real risks of the ride getting pretty rough.

All these projections inform our courses of action.

Nate's question boils down to, how certain would we need to be of gloomier projections to change our plans?
Or analogously, how steep will it get before we change gear?
Perhaps we don't have too many gears; maybe just

1. sell up and book round-world hiking trip
2. work, save, skill-up
3. cower in bunker

- as a set of routines to stick to.
Outside of routines, there's reactions to events that can't be dealt with by routine; storm, flood, war, etc.

So in summary, I think we'll mostly stick to a middle-outcome-appropriate routine until we have something intrusive to react to.

0.02; j