Originally a comment by James Garnett on Triple threat.

I’m not a climate scientist, but my professional work is in control theoretic modeling of real, nonlinear systems. It’s difficult to underscore just how frightening a positive feedback runaway can be; the results are catastrophic for the system in question, even for small things. For example, induction motors form the basis of modern processing engineering, so their function and use are taught in most electrical engineering programs. One aspect of these motors is that when they are running, if one opens the field circuit while the driving circuit is still energized, the torque of the rotor becomes infinite (in theory), because the field acts as a negative feedback control mechanism.

Some years ago I witnessed what this really means when a student did exactly this to a half-ton induction motor bolted to the floor of a lab with one-inch bolts: in a matter of seconds it tore itself free of its anchors and proceeded to fly around the room, wreaking devastation to every inch of it; it was only by sheer luck that nobody was hurt or killed.

Now scale that scene in your mind up to the entire planetary climate: that’s what we’re facing. This is why climate scientists often seem to use such hyperbolic language, but what is frustrating is that due to the nonlinear nature of the overall system they cannot make accurate predictions of when this will happen. All we can really say about such systems is that when the bad stuff truly gets going, it’s going to get really bad, really fast.