Euthenising the Universe
This quirky and disturbing preprint (by a couple of astrophysicists with impeccable credentials) has been doing the rounds over the past week or so, and if I’m reading the commentaries right it’s taking the Copenhagen Interpretation of quantum physics to its logical conclusion — specifically, that whole Schrodinger’s Cat thing that says nothing actually exists until an act of observation collapses the probability wave and forces the universe to make up your fucking mind already. If you buy this interpretation, then a bunch of astronomers who looked at a supernova back in 1998 may have — by that very act of observation — shortened the lifespan of the whole universe. (The obvious question about whether the universe-altering observations have to be made by human astronomers — or even humans for that matter, given that at least a few photons from that supernova must have fallen onto the retinas of everything from cats to Cardassians long before now — was never addressed.)
I read this paper. More precisely, I ran my glazing eyes over each line and each equation in turn, while moving my lips. And even though I kinda recognised some of the Fourier transform stuff, it was pretty much all over my head.So I showed it to a biochemist I occasionally hang with; she got all squee-y because she kinda recognized the imaginary-numbers stuff, but she wasn’t much help beyond that. Fortunately we happened to be in a bar with a couple of astrophysicists, who had been roped into this community outreach program where experts on various subjects fend off questions hurled at them by drunken patrons. One of these experts actually specialized in the whole dark-energy thing; the other was a former student of his. So I hurled this weird Krauss-and-Dent paper at them, and this is what what they said:
The Master said that the paper took quantum theory to its “logical extreme”, and then kind of shrugged and said “But what are you going to do? It’s not like we’re going to stop looking.” He also allowed that the whole thing sounded kind of like worrying that the elephant that supported the world was going to fall off the back of the turtle that supported the elephant.
His apprentice said “If in fact the astrophysics community has shortened the lifespan of the universe, I’d like to take this opportunity on behalf of Canadian astrophysicists to be the first to apologize.” I liked that.
But neither of them said the paper was wrong. Neither pointed out any sort of fundamental error in the math or the conclusions. In fact Carlberg, for all his grousing about giant turtles, did grudgingly concede that the conclusions followed as a “logical extreme” of the theory. I find this disturbing.
Of course, the Copenhagen Interpretation does have competition. There’s also the Many-Worlds Model, which in contrast to the nothing-is-real view, claims that everything is — that there is no probability wave, only an endlessly-proliferating infinity of parallel universes that spawn wholesale every time an electron has a choice between flipping this way or that. This theory also carries some profoundly ugly implications (it confers credibility onto Sliders, for one thing; also, nobody has explained to me where the extra mass for all these universes is supposed to come from), but it seems to be gaining ground amongst the theorists.
Still. Just to be on the safe side, it couldn’t hurt if we all agreed to walk around for a while with our eyes closed. It might buy us some time.
Update 1715: OK, looks like a false alarm. Initial popsci reports were all causal-this and shortening-the-lifespan-of-the-universe that, but as AR has been kind enough to point out, Krauss is actually quoted in the article I linked to as saying he didn’t mean to imply causality. Move along. Nothing to see in the comments. (Unless you want to see AR pointing out how credulous I can be…)
A bit like Greg Egan’s Quarantine it seems.
As a physicist myself, I’m just not getting the same thing out of this as you seem to be. It looks to me like it’s saying that the observations imply that the universe is in a particular quantum state; not that the act of observation itself caused that state. In fact, the commentary you linked quotes a saying exactly that.
If anything, observation would serve to delay any phase-transitions the universe has yet to undergo, as an unstable quantum system that is continuously observed will never decay, a principle known as the Quantum Zeno effect.
I should also point out that the popular conception of what constitutes “observation” in quantum mechanics is often poorly understood. Conscious observers are not favored, nor are they they only things capable of causing collapse. We are, after all, part of the universe, and if we are able to cause a collapse, then so is the rest of the stuff in the universe. It is possible to exploit this fact in order to do such things as gain information about a sample without interacting with it, as is shown in the Elitzur-Vaidman bomb-tester experiment.
Okay then, what am I missing here? Lines like “Have we hastened the demise of the universe by looking at it?” and “we may have accidentally nudged the universe closer to its death by observing dark energy” sure look like they’re talking about causality to me. Now that I reread the Telegraph article, I do recognize that disclaimer from Krauss — “I did not mean to imply causality – namely that our measurement itself reduces the lifetime of the universe”. But then the same bloody article says “By observing the system … we may pick out one single ‘quantum state’, and therefore force the system to change its configuration.” That sounds pretty causal to me, especially in light of a headline that reads “Mankind ‘Shortening the Universe’s Life”—
Actually, you know, reading over this again, it looks almost as if Krauss is doing his damndest to distance himself from the spin that New Scientist put on his work (it was NS that really went to town on the causality angle). Whereas other sources I’ve read claim that this act of observation “reset the clock”, the Telegraph article has Krauss claiming only that “our detection of the dark energy may provide evidence that the universe will ultimately decay” and “we may have confirmed that we are…”
Ah. I see it now. Got all carried away by the headlines and the original article. Didn’t pay enough attention to the disclaimers in the followup.
Thanks for setting me straight on this, AR. The rest of you: never mind…
Good call gioppe – when this article showed up on Drudge, I too thought this was the reclusive Australian’s novel come true, albeit without the shutters around the solar system.
“Ah. I see it now.“
That it took so long is undoubtedly because you were walking around with your eyes closed. Just don’t ask me which was the cause, and which the effect.
John Baez wrote a post on this very topic December the first…
http://golem.ph.utexas.edu/category/2007/12/astronomers_blamed_for_death_o.html
… taking the various news services to task for being a wee bit sensational and (far more importantly) explaining the paper in terms which I can pretend to mostly understand.
The part that really stumps me is that Baez asks this question:
“has anyone ever seen deviations from exponential decay for radioactive nuclei or similar systems?”
… and he’s already mentioned the quantum zeno effect. So that doesn’t count. Why doesn’t it count?
Help, please?
That link didn’t come out right, did it? It’s at http://golem.ph.utexas.edu/category/ under “Astronomers Destroy Universe”
re the Many-World hypothesis: i read a fascinating time-travel book a few months back which beautifully neatly eliminated all the paradox stuff by positing that the whole branching idea ran _backwards_. ie, that the universe’s probability wave was collapsing forward in time to a single certainty, rather than expanding forward in time.
the sheer elegance of the consequences was magnificent.
i just wish i could bloody remember which book/author it was.
iirc, it was written in the 50s or 60s or thereabouts, by one of that tiny band of sheer geniuses that sold far far less than they should have. don’t think it was sladek — might have been kornbluth?
Sal said…
re the Many-World hypothesis: i read a fascinating time-travel book a few months back which beautifully neatly eliminated all the paradox stuff by positing that the whole branching idea ran _backwards_
That does sound like a cool twist, although I don’t see what paradox it resolves — doesn’t the many-worlds interpretation resolve causality paradoxes even if the branching goes forward?
For that matter, one thing I’ve always wondered about with this whole parallel-universe shtick; where does all the necessary mass come from to create these infinitely-proliferating alters? Anyone know? AR, you there?
My hand wave explanation is that if a branch occurs with 20% probability to universe A and 80% probability to universe B, then 20% of the original mass winds up in universe A and 80% of the original mass goes to universe B. This idea has problems, not sure if they are serious.
You could probably build a toy model with two interacting particles. Pretend that one is the “observer” and the other is the system.
I thought the whole point of the many-worlds interpretation was that it avoided the “probabilities” issue entirely; every possible alternative happens with P=1.0, which necessitates the proliferation of universes to accommodate the existence of all those alternatives. And I don’t see how you could just divvy up the mass of the parent universe between daughter parallels; given the rate of proliferation, you’d end up with, like, half an atom per universe in about two femtoseconds, yes?
Or am I missing something?
Nngg. I wrote in an unclear way.
When I say “probability” I mean in the classical sense. So you flip a coin. Two different timelines. 50% of your mass one way, 50% the other, 100% of the time.
The observer and system lose mass, but they all lose the same amount down to the electrons, so who notices? Mass is relative, after all.
And yeah, in an instant each universe is down to half an atom’s (previous) mass. If mass is continuously divisible then there is “plenty of room at the bottom” and this can go on forever. But is it? …so I tried the toy model. Solving involves tensors. AR, help?
These could be big problems. But what’s the alternative? Observation (conscious or not) collapses the wavefunction? That way lie silly questions. Many-worlds at least is deterministic.
Question 37 of the Everett Interpretation FAQ by Michael Clive Price: Building a mind that can split when the world does not, allowing limited communication between futures.
http://www.hedweb.com/manworld.htm#detect involves building a reversible quantum computer.
Keith said…
These could be big problems. But what’s the alternative? Observation (conscious or not) collapses the wavefunction? That way lie silly questions. Many-worlds at least is deterministic.
I was having beers with a couple of guys whose education in such things is far more recent than mine, and they tell me I got the whole model wrong: that many-worlds does not entail infinite proliferation of universes at each decision point, but rather a simultaneous proliferation of all possibilities at the beginning of time. So the timelines extend from t=0 like spokes on a wheel, not (as I’d thought) like branches on a shrub. So there’s no ongoing splitting, and thus no loss of mass— every single possible timeline is realised up front, and proceeds in isolation unto heat death. The quantum uncertainty we experience is an illusion, since until we make a specific observation we don’t know for sure which timeline we happen to be in.
Have I got that right? Or at least, righter?
Question 37 of the Everett Interpretation FAQ by Michael Clive Price: Building a mind that can split when the world does not, allowing limited communication between futures.
That is wild. I’d had no idea that such tests were even theoretically within grasp. I’ve bookmarked that whole faq to go back and study more extensively, some time when I have six weeks to kill…
I was having beers with a couple of guys whose education in such things is far more recent than mine, and they tell me I got the whole model wrong: that many-worlds does not entail infinite proliferation of universes at each decision point, but rather a simultaneous proliferation of all possibilities at the beginning of time. So the timelines extend from t=0 like spokes on a wheel, not (as I’d thought) like branches on a shrub. So there’s no ongoing splitting, and thus no loss of mass— every single possible timeline is realised up front, and proceeds in isolation unto heat death. The quantum uncertainty we experience is an illusion, since until we make a specific observation we don’t know for sure which timeline we happen to be in.
So the (already separate) timelines differentiate apart gradually as time goes by. That sounds a lot like a “hidden-variable” interpretation – the idea that just because certain things are unmeasurable doesn’t mean they are undetermined. It avoids that awkward moment where the universe splits, but you wind up having unobservables. Sounds cautiously like a win…
Have I got that right? Or at least, righter?
Better to ask your drinking buddies! I’ve only had three semesters of quantum. All I can say is no problems immediately jump out at me. Heard rumors that the hidden-variable theories are all nonlocal but I’m not good enough to do the math myself yet, and I’m also not sure that nonlocality is a problem anyway.
By nonlocality I mean the notion that particles communicate their hidden states with each other in a superluminal fashion. Aka spooky-action-at-a-distance, EPR paradox, Bell inequality, “the fundamental interconnectedness of all things” as Douglas Adams would say.
That is wild. I’d had no idea that such tests were even theoretically within grasp. I’ve bookmarked that whole faq to go back and study more extensively, some time when I have six weeks to kill…
Glad you enjoyed it! Exciting, how far we’ve come. I like the faq for being math-light and accessible. No friggin tensors. A splash of thermodynamics just for fun.
As for using AI to ping nearby universes… Maybe we don’t have to wait that long. It seems like a bit of overkill, doesn’t it?
Quantum computers (if I read the papers right and I’m not sure I did) use parallel computation across nearby universes. Maybe *your* computer can’t do the job, but if it has 255-tuplets it can talk to for a nanosecond, the team can solve the problem.
Noise always knocks the computer back into a classical state in the end, of course… but if it stays superposed for long enough, maybe we can think of an experiment to detect the other tuples. Some prediction that Everett makes that Copenhagen disagrees. Not exactly communication, just a proof that there’s something there. Heck, I bet we could do that in only 25 years.