Which is the hingy-est century?

A while back I linked to Holden Karnofsky’s argument that forthcoming times are likely to be the most important for determining the course of subsequent history, or “hingy-est” of all time.  So I thought I should address the issue directly myself.

In my view the greatest danger to civilization is war, rather than AGI.  Rather than rehashing that debate (see Holden’s view here), let’s just take the war view as given and see where it leads.

I see a few distinct possibilities:

1. The relatively peaceful world order since WWII will continue for the indefinite future, albeit with ongoing evolutions and modifications.  If that is true then the second half of the twentieth century might be the hingy-est time because that was when we built enduring peace.

1b. But the postwar era doesn’t have to be the hingy-ist time under that view.  It might be that “the finding of peace” was highly likely or inevitable, sooner or later.  Maybe it was the Industrial Revolution that was more contingent, and without that we would have found ongoing peace but at much lower living standards.  In that case the British seventeenth and eighteenth centuries could well be the hingy-est time.

But sadly, while I see #1 and #1b as possibly true, they are not for me the most likely scenarios.  There is also:

2. Humanity will fight a very destructive war at some point.  It will not kill everyone but it will slaughter a significant portion of the earth’s population and put the rest into something like “African living standards plus Balkans governance.”  With no turnaround in sight, if only because it is so hard to cast off those institutions once they are in place.  Protection against subsequent existential risks will be harder as well.

In that case the hingy-est time or century would be whenever that war comes, or whenever some set of preconditions made such a war inevitable.

To be clear, I think the chance of such a war is quite low in any given year.  You don’t need to be shorting the market.  Still, if you let the clock tick long enough, such a war is bound to come.

Now is the next 50-100 years the most likely era for such a war to arrive?  I don’t see a strong argument why we should have such a definite intuition here.  We’ve had some version of MAD with nuclear weapons for quite a few decades, and it has mostly worked out OK.  At some point upping the firepower might shift that balance (drone assassins of political leaders?  Or something that comes 137 years from now?).  I don’t find it easy to have good intuitions on this question.

I am reminded of my earlier post on how long it took the NBA to truly adopt and exploit the logic of the three-point shot.

Even introducing strong AI doesn’t settle it for me.  Strong AI might lengthen the reign of (relative) peace, rather than shortening it.

Many things, both positive and destructive, can take longer than you think.  And as a general reminder, foreign policy outcomes are extremely difficult to predict, even across a small number of years much less decades.

So I don’t know when the hingy-est century or era is likely to be.