Trespassing on Einstein's Lawn (46 page)

“The global approach has to be wrong at some level,” he told us once we had settled into our seats.

That much was evident even without horizon complementarity—just look at the measure problem. In a global multiverse, probabilities drown in infinities and physicists lose the ability to make scientific predictions—which was, after all, the point of science. Bousso suspected that those infinities were related to the duplicate information you get when you try to view the world from inside and outside a horizon.

“The lesson to take away from the black hole information-loss paradox is that you have to restrict to the viewpoint of only one observer,” he said. “It seems very natural that this is a restriction we should impose on the multiverse.”

So he did. And while he was at it, he looked to see if such a restriction would solve the measure problem.

“I thought, why not do something that we've been forced to do by the black hole information-loss paradox anyway and kill two birds with one stone?” Bousso said.

The usual method for coming up with a measure went something like this. Arbitrarily choose to call one dimension of the multiverse “time” and the others “space,” despite the fact that time and space are observer-dependent. Look at a finite region of the multiverse at a given time and count, say, the number of universes that have a certain value of dark energy relative to the total number of universes in the sample. Take the limit of that ratio as the sample size goes to infinity. Then compare your answer with the one universe we actually see. If your
measure says that the probability of a universe with a value of dark energy as small as ours is basically zero, trash it and try again.

Bousso, however, took a completely different approach: only count what's inside a single observer's light cone, otherwise known as a causal patch. That way, you don't have to come up with suspect ways to whittle down infinity because the whole scenario is finite to start. What's more, you don't have to worry about overcounting the redundancies that crop up when you cut across horizons.

But I was confused. “If you're looking at a single causal patch, why do you have to deal with probabilities at all?” I asked. If you're restricting yourself to a region of spacetime in which everything you'd ever want to observe is within your causal reach, what do you need a measure for? You don't have to calculate the odds of seeing something—just open your eyes and look around.

Unfortunately, Bousso said, it's not so simple. In a universe ruled by eternal inflation, the physics within a causal patch changes over time, since every vacuum with a positive value of dark energy is unstable and subject to decay. According to the standard story, our universe started off in a false vacuum—a temporarily stable state that is not the lowest possible energy—and then plummeted downward, inflating faster than light in the blink of an eye, until it hit the true vacuum below. That's where we've been living for the past 13.7 billion years. But as we now know, our vacuum wasn't the true vacuum, either, even if it was slightly less false. Our vacuum is permeated with a small but significant amount of dark energy, whereas the true vacuum would have no dark energy at all. Our universe is perched atop a stable plateau, but the right quantum push could send it careening over the edge and soaring down toward rock bottom, stopping at various lookout points along the way. Each stop is a big bang, each plateau a universe. This finite but wildly varied plunge through cosmic history is a kind of multiverse in its own right, despite the fact that it never refers to any reality beyond the light cone's edge.

Just like the decay of a radioactive atom, the vacuum decay of a given universe is random, ruled by nothing more certain than probability. As an observer, you don't know which universe you're in, so if you
want to make testable predictions, you need a probability measure to calculate what you're most likely to see.

“So you start with some vacuum, you calculate all its probabilities for decay, and you add them all up,” Bousso said. “Instead of a global multiverse you have an ensemble of vacua within a single causal patch.” If you want to predict something like the value of dark energy, you don't have to look to other bubbles in the global multiverse; you just have to consider the possible histories within your own cosmic horizon. The result was what Bousso named the causal patch measure.

It seemed like a good plan: take Occam's razor to everything outside your horizon, make spacetime finite, and generate a probability measure based on a history of vacuum decay, harnessing the anthropic power of the multiverse without having to deal with the Popperazzi or the metaphysical burden of eternally inaccessible universes.

Then it got even better.

“I thought I was going off in a radically new direction by rejecting the global viewpoint,” he said. “Then I learned that the local causal patch measure reproduces the exact same probabilities that you get with the global light cone measure. That came as an enormous shock to me.”

The global light cone measure was more “global” than “light cone.” It was one of many global measures on the market, the kind that counts across horizons, the kind that, as Bousso said, has to be wrong at some level. This particular global measure, however, had something going for it that the others didn't: it happened to give the exact same probabilities as the causal patch measure.

“I had no idea that two measures could look so different and turn out to be the same,” Bousso said. “That was amazing to me.”

The two views—the God's-eye view of the whole multiverse and the internal view of a single observer—were completely equivalent. Dual. So even though the global view is fundamentally flawed, the local causal patch measure offers us a way to still talk about a global view as if it actually means something. To talk about a multiverse without ever referencing anything beyond our own horizons. I jotted in my notebook:
Best of both worlds.

“We had been thinking there was a kind of shift under way in cosmology,”
I said, “from a God's-eye view of the universe to the point of view of a single observer.”

“Not everyone sees it this way,” Bousso said. “But I would argue that it's inevitable. Lenny [Susskind] saw that in the context of black holes you get the wrong results when you think of the bird's-eye view of inside and outside the horizon—it looks like you're Xeroxing information and violating quantum mechanics. My view was that it was too radical a statement to restrict just to black holes. It must be telling us something deep about how things work in general.”

“But you showed this duality between the global and local pictures,” I said. “Doesn't that make the global view valid again?”

“The local view is still more fundamental,” Bousso said. “It has to be. The idea is to somehow build a global multiverse out of these local patches.”

I nodded. Maybe there was hope after all. Maybe you could take multiple sole authors and patch them together, though I had no idea what that book would look like. Would it alternate authors from one chapter to the next?

Sharing a laugh with Raphael Bousso, UC Berkeley
W. Gefter

“Would you ascribe reality to things outside our cosmic horizon?” I asked, echoing Susskind's deepest question in cosmology.

“I'd say there are things happening there that might be a tiny bit different from what's happening here, but I would not talk about them simultaneously. So it depends on how ‘real' you want things to be,” Bousso said with a laugh. “It is certainly my belief that the most fundamental way to describe cosmology is to restrict to one causal patch.”

“We're trying to figure out what the ingredients of ultimate reality might be,” I said. “And clearly we need quantum gravity to find them. What could they be? Are they strings?”

I was pretty sure I knew the answer to that, since AdS/CFT had knocked strings off the IHOP list, revealing them to be holographic projections of ordinary particles, neither description more “real” than the other. Still, Bousso was a string theorist, so I figured I'd ask.

“It's one of the big questions,” Bousso said. “String theory wasn't discovered in the way that theories are usually discovered in physics, where you have some idea of the principles and the basic ingredients and you build it from there. We just sort of stumbled upon string theory, and through mathematical consistency we keep discovering new ingredients. As a result, we don't know which ingredients are fundamental. In one setting you may want to think of strings as fundamental, and in another you might think of D-branes as fundamental. It's not even clear that there's a unique answer to the question.”

I quickly scribbled in my notebook:
D-branes?

If strings and particles were merely two ways of looking at the same thing, I thought, what's the thing? As an ontic structural realist, I figured the thing had to be the holographic principle itself. But what was that telling us? If a holographic principle,
why
a holographic principle?

“Is the holographic principle the key clue to quantum gravity?” I asked.

“It shows a relationship between geometry and information that's completely general but of which we don't understand the origin,” Bousso said. “We can say what the relationship is and we can look at nature and check over and over that it really holds true, but it's a bit like a conspiracy. It's got to have a deeper reason. And we think that reason has something to do with quantum gravity, or more precisely with a unified theory of matter and quantum gravity. On one side of the relation you have geometry, the area of surfaces in spacetime, and of course spacetime is gravity. On the other side you have the amount of information that can fit in adjacent spacetime regions, and information is really just the number of quantum states. So you have gravity on one
side related to quantum theory on the other. The relationship is so universal there has to be a simple reason for it. It's like, why do things fall at the same rate? The hope is that the holographic principle can play a role analogous to the role the equivalence principle played in guiding Einstein toward general relativity.”

“The fact that you get the wrong answers when you use a global, God's-eye view,” I said, “what is that telling us about reality?”

Bousso thought for a moment. “There's a sense in which it's telling us that what we call reality, no matter what it ends up being, is probably an approximate concept. If a typical observer has access to a finite number of quantum states, because your past light cone has a maximum area that dictates how much information will fit in it, that means there's a limit to how sharply you can measure anything, how precisely you can describe the world. There's clearly no infinitely sharp sense in which the world is some particular way.”

On the train ride back to San Francisco, I thought about what Bousso had told us. The local view—the perspective of a single observer, a light cone—has to be fundamental. If you want a multiverse, you can try to patch it together out of local views. But at the end of the day, it's an illusion. We could officially cross the multiverse off the ultimate reality list.

I tried to imagine what a multiverse cobbled together from local views would look like, but my brain couldn't piece it together. I knew it couldn't be the sea of bubble universes that cosmologists always show to illustrate the thing. Such a picture assumes that the bubbles have some definite, observer-independent existence. But it's not like the edges delineating individual bubbles are physical walls floating in some invariant space—they are internal markers of the boundary of a given observer's point of view. Viewing the multiverse from the outside would be like seeing every possible perspective simultaneously.

Whatever that would mean.

I couldn't help thinking of Borges's Aleph. In the story “The Aleph,” the protagonist discovers that a man named Carlos Argentino Daneri
has found, under his basement stairs, an Aleph: a point in spacetime from which every point can be seen simultaneously, from all angles. A God's-eye view.

“The Aleph was probably two or three centimeters in diameter, but universal space was contained inside it, with no diminution in size,” Borges wrote.