December 29, 2010
3.7 Billion Years To Last Restaurant At End Of Universe?
At UC Berkeley Raphael Bousso and some friends think the universe has to be of finite duration. Time to book a reservation.
There is a 50 per cent chance that time will end within the next 3.7 billion years, according to a new model of the universe
About that 5 billion year birthday party some of you have planned: Do you feel lucky?
An infinite universe is crazy because extremely low probability events can happen an infinite number of times?
Their argument is deceptively simple and surprisingly powerful. Here's how it goes. If the universe lasts forever, then any event that can happen, will happen, no matter how unlikely. In fact, this event will happen an infinite number of times.
Just how many problematic extremely low probability events are there? Why wouldn't their probability drop to 0 at some point as the universe becomes too spread out?
From the abstract:
Present treatments of eternal inflation regulate infinities by imposing a geometric cutoff. We point out that some matter systems reach the cutoff in finite time. This implies a nonzero probability for a novel type of catastrophe. According to the most successful measure proposals, our galaxy is likely to encounter the cutoff within the next 5 billion years.
You can read the full PDF. One weirdness: if you ever find yourself falling toward a black hole time might end before you reach it. Not sure if that's a feature. They also discuss whether time ends sharply or in a smear.
Lift up your eyes to the heavens,
And look upon the earth beneath;
For the heavens shall vanish away like smoke,
And the earth shall wear out like a garment,
And its inhabitants shall die with them; but
My salvation shall be for ever, and
My righteousness will never fail.
Oh When the sun refuse to shine.
Lord, I want to be in their number,
When the sun refuse to shine.
I've read that before, and regard the reasoning as quite lame. It's a philosophical cousin to Zeno's paradox, if Zeno had actually been so silly as to think Achilles really couldn't pass the tortoise. They're confusing a limitation of their own mathematics with an actual physical principle. Just because our current mathematics of probability doesn't handle infinite samples sizes well, doesn't mean infinities can't exist.
For this reason, I suspect the whole paper is something of a farce, not really meant to be taken seriously.
Slighlty off-topic: but can anyone explain to me how the Hindu cosmology system got so much about the universe right? i mean,they were thinking in terms of trillions and billions years for the age of the universe when the Vedas were written thousands of years ago!
I'm not a physicist, but I was thinking the same thing as Bellmore. Why should we assume a priori that our model of the universe is indeed universal?
Here's an analogy: Suppose you are an ancient man running around the jungles of Central Africa. In your brain, there will be a model of the universe as you know it: There is the ground; there are plants and trees growing; if you cut down a plant it will grow back; the day is always about the same length as the night; and you can drink from open bodies of water.
Now suppose someone suggests to you that there are huge deserts where nothing grows; that there are huge bodies of water which are too salty to drink; and that there are places where the day is much longer than the night for part of the year and vice versa at other times.
You might be tempted to argue that the suggestion can't possibly be correct -- that it violates the fundamental principles that (1) stuff always grows from the Earth; (2) bodies of water are always drinkable; and (3) days are always roughly the same length as nights. But if you did, you would be making a mistake. You would be confusing your model with reality.
How is the argument in this article any different? Again, I am not a physicist but it does seem suspect to me.
Well, that, too. But what I really meant is that their entire argument boils down to, "My math doesn't handle a universe that lasts forever. Therefore the universe doesn't last forever. It would make my math easiest if the universe lasted about *this* long. Therefore it's going to end after that amount of time." "there is no well-defined probability distribution without the cutoff;"
But why should the universe care if he can't define a probability distribution without assuming a cutoff? Wake me when he comes up with a mechanism, and demonstrates evidence of it. All he's demonstrated so far is the limitations of his statistics.
Mathematicians have been rigorously reasoning about completed infinities ever since Cantor.
Some of us have even managed to stay sane while doing so.
Nobody can hazard a guess at how the ancient aryans could get the time-scale of the universe so near the present day one,thousands of years before they had radioactive date to measure the age of rocks or microwave radiation data from space? Thanks.