repent, for the end of time is relatively near!

According to four physicist time will keep going for billions of years until one day it will just... stop.

Or at least so says a quartet of physicists who are certain that the flow of time has a 50% chance of ending in about 3.3 billion years because according to them, in a universe that expands infinitely, every unlikely event, as far as physics is concerned, will happen an infinite amount of times. Therefore, they continue, to preserve the laws of physics, time must end so all these bizarre things don't happen and then keep on happening by sheer probability.

Yeah, I know, I'm sort of having trouble following their logic myself. They seem to be saying that the flow of time must stop so weird things don't happen, and then use some bizarre formula to come up with what most of us would call a fairly random 50% chance of this happening in 3.3 billion years, and that in 5.3 billion years, it will grind to a proper, complete halt. Yeah, about that… To say that I'm not convinced is to put it mildly.

Time, as we understand it today, is a function of entropy. When you're talking about time ceasing, you imply an end to entropy as well. Without the flow of time, how will galaxies keep moving, planets keep orbiting, or stars keep producing energy? And how would a universe which still has enough fuel for stars and planets for over a quadrillion years suddenly grind to a halt? The paper's authors don't indulge in an explanation, they simply trot out a time-space metric and go from there, following a disorienting thought experiment with watches and time horizons.

Again, their only justification for having time end is to prevent really improbable physical events from manifesting themselves because if the universe's expansion continues forever, then every single conceivable event, no matter how bizarre or improbable, will happen an infinite number of times simply because it has the luxury of having infinite chances to happen. And while the paper's authors may view this as a problem, I really don't understand the big deal. It's as if the authors decided that if physics can't predict what the universe could do when it's in its "hey you kids get off my lawn!" stage, they might as well pitch the whole thing.

Sure, with an infinite expansion, anything can and will happen, but so what? We would still be able to predict physical phenomena and their behaviors in local space and physics as we know them will still have meaning and purpose. We've talked about physicists looking to simplify cosmology before, but this isn't simplifying a complex and growingly over-elaborate sub-fields and reconciling them with what we already understand. This is mathematically ending the universe because something odd is sure to happen in the far future.

As you can probably guess, this paper hasn't been peer-reviewed and the only reason it got the buzz it did what thanks to the Technology Review arXiv blog which is now known for its glowing write-ups of statistical numerology as applied to space and time, lazy quantum hypotheses of multiverses and black-hole driven Big Bangs, and supposedly viable relativistic hyper-drives which are anything but. Really, I know the peer-review system isn't perfect and even listed some of its spectacular failings in recent years, but it doesn't mean that we need to start going around it and keep flooding popular science sites with cryptic and poorly thought out papers.

See: Bousso, R., et al. (2010). Eternal inflation predicts that time will end, arXiv: 1009.4698v1