The Physics of Everything

Over the past century, physics has pushed our understanding to the distant edges of the universe and to its very first moments. "It's just amazing how much has been learned," said Eva Silverstein. She noted that the extraordinary uniformity of the cosmic microwave background radiation leftover from the early universe is now attributed to an early "inflationary" period driven by a new quantum variable. Unfortunately, she said, "that variable can't help but have quantum fluctuations. It's those fluctuations that we think seeded the structure that we see," such as galaxies, as well as the tiny deviations from uniformity of the microwaves. "But at the same time, those fluctuations mean that that all we see is one realization," essentially a cosmic roll of the dice that determined the specific structure we live in.

Using the cosmic background and other data, said Neil Weiner, "we can determine the constituents of the universe very precisely: 4% is atoms, about a quarter is dark matter. and the balance is dark energy." Dark energy's uniform effects leave few clues as to its nature, but dark matter is more promising because it reveals its clumpy distribution through its gravitation. Still, "how can you look for it if you don't know what it is?" Weiner wondered. "I don't know that we'll ever know."

The traditional view of black holes was that they left essentially no clues to what had fallen in. These days, however, said Vijay Balasubramanian, "the dominant view is that there is a mistake in the reasoning concluding that information is lost." However, this information is not in a form that would be useful to an observer, he said. "It's an in-principle question." [For further discussion on this topic, see Lecture 6.]

Human limitations

For some modern physics theories, such as string theory, it is hard to envision experimental tests, raising the question of whether experimental validation should be required. It is hard to answer this question in the abstract, Weiner said. He cited the recent evidence for the Higgs boson, which had been predicted a half century earlier. If the experiment had not been done, "would you say you know there's a Higgs boson?" he asked. "I'd probably say no."

A similar problem arises for the "multiverse": multiple non-communicating universes that are predicted by many inflation models. "An optimistic scenario would be that we test our theory of physics that we can observe locally," Silverstein said, which would then inspire confidence if "that theory has the further consequence of a multiverse. We're not there yet." [For a philosophical perspective, see Lecture 2.]

"We're not going to measure everything," Balasubramanian noted. "The scientific method is making some small number of measurements, making a relatively simple model that we can contain within our heads, and then betting that works over a much larger range." The question is "whether that kind of extrapolation has limits," he said.

"We get most excited when it's about principles," Silverstein said. Balasubramanian agreed, citing the Jorge Luis Borges story "Funes the Memorious," about a boy who knows every fact but cannot truly think. In saying "to know," he said, "we don't actually mean knowing all these things separately. We mean a sort of categorical knowledge of the principles," he said.

"This is discounting the wonderful things that you can see in phenomenology, which you would not have seen as a principle," Weiner objected. "There's the laws that maybe we hope that we can get to from a finite set of experiments," he said, "and then there are some questions about what is actually out there. There are different types of limits for these different types of questions."

Our brain architecture could also limit our comprehension. "A cat can't understand calculus," Balasubramanian said. "There's no a priori reason why the human brain is capable of producing the representations of knowledge necessary to understand the universe." The fate of information in black holes, said Silverstein, is an example "where we are currently limited by our lack of intelligence."

Known unknowables?

In developing theories, "the challenge is that you can make many more-complex theories that will fit the limited data," said Balasubramanian. "So we use Occam's razor [which favors the simplest workable theory] as a rule of thumb to guide us through this maze. I'm constantly surprised it works to the extent that it does."

This approach can fail, though cautioned Weiner. "If you stick to this idea of simplicity then sometimes you can miss" an accurate picture. As an example, he imagined that we lived in the dark matter sector and tried to come up with a model for ordinary matter. "You would never come up with the description of the universe that actually is. You would insist on something which is very, very simple," he said. "We're going to describe it as simply as we can, but that doesn't mean that's actually right."

In spite of these challenges, the three theorists saw no benefit to invoking a deity or something else beyond physics. "I don't know really what 'going outside physics' means," Silverstein said. Religion seems to fill a deep psychological need for some people, Balasubramanian said, but "I don't think it's useful, at least to me, as a discovery method in science."

"Actually it could be dangerous," Weiner said. "Even if it ends up being wrong that we can explain things by the scientific method or by physical principles, you should at least pursue it as if you can, because if you don't then you run the risk of missing out on really tremendous stuff."

"We have physicists, who profess to have no interest in philosophy, actually doing quite a lot of philosophy, often without knowing it," said author Jim Holt. He cited Steven Weinberg's dismissal of positivism [which demands that physical theories refer to observable things] as "dangerous," while Stephen Hawking "proudly calls himself a positivist." These two highly eminent physicists "seem to have a philosophical disagreement," Holt observed. "The question is, are they're doing it well or poorly?"

Recent contributions of philosophy to physics, said David Z. Albert, include the question of "why is it that measurements have determinant outcomes?" In that instance, "physicists were not at all surprising us with the range of questions that they could answer. They were surprising us in a different way: with the narrowness of the questions they were willing to consider." Philosophers showed that "there was a genuine scientific question," he said. "This I think was very productive."

Need for interpretation

"Physicists can be surprisingly good at answering difficult questions," said Hans Halvorson. "Philosophy's role is not to police physics but to make sense of it." As an example he mentioned the use of the word "particle." In the current theory, "there are quantum fields, but there are not really particles," he said. Philosophers "had the luxury to sit back and say, I wonder if there really are particles?"

Albert discussed the interpretation of time, noting that physics has long "spatialized" time, treating it as a "fourth component of the address of an event." By this gesture, he said, physics "cut itself off from the possibility of confronting what was really interesting about the temporal," so that any discussion of time "flowing" sounds like "meaningless gibberish." However, "our experience of the temporal coordinate is very, very different from our experience of the spatial coordinates," he said, and "physics needs to give an account of how the way that mass and energy is distributed gives rise to this."

Quantum mechanics, Albert said, has threatened the long-standing aspiration in the West "to achieve some kind of understanding of the world as a separate, real, free-standing object," with properties independent of whether anybody is looking. This vision came to be regarded "as naïve, as presumptuous, and at any rate as false." But recently, he said, philosophers and others have shown that this conclusion was "reached too fast, that the original aspiration of producing an understandable, realistic account of what was going on, rather than just a prescription for predicting the outcomes of experiments, is now widely considered much more alive and well than it had been through most of the previous century."

Because observations are central in quantum mechanics, "physicists are much more willing to talk about consciousness as an integral part of physical processes," said Holt. In contrast, "contemporary philosophers tend to be much more realistic, in the metaphysical sense."

Goals of physics

Philosophers are also well equipped to examine the nature of scientific deductions, the role of experiments, and the significance of models. In Niels Bohr's framing of quantum mechanics, for example, "science is no longer trying to describe an objective external world," Halvorson said, but simply to predict the outcome of experiments. "Philosophers were very instrumental in saying the task of science can't be put so simply."

A more recent challenge comes from the independent universes that naturally emerge in some cosmological theories. "The multiverse often gets accused a little too quickly of being something unverifiable," Albert said. "It may be in some principled way, not possible to, say, directly look at the other universes, but that doesn't mean that we won't come to have good empirical reasons of all sorts of other kinds to believe these theories." If other predictions are verified and increase our confidence in the multiverse framework, "it doesn't seem a departure from traditional scientific standards of confirmation."

Still, Holt noted that the predicted universes could have completely different laws of physics, different fundamental constants, and even different numbers of spatial dimensions. "Suddenly the ability of physics to predict things is greatly eroded, because things that looked like they're fundamental parts of our universe—the constants that define the standard model, for example—become contingent, like local weather." For some people, "that's a cheap way of explaining too much, and it's not in the best tradition of science."

An ongoing challenge for collaboration between physics and philosophy, Halvorson said, is that "people become tagged by their departments. These boundaries should be more fluid than they are." In spite of recent public criticisms, physicists' defensiveness about fundamental questions "is on the decline," Albert said. "One of the chief engines of intellectual progress is that people die. Young people find it a more natural extension of what they are doing."

"In physics it's very important to be able to separate time scales or length scales," Chazelle said. "if you move to a different scale, you have a different language that allows you to completely ignore what happens below. In biology this rarely happens."

In spite of the difficulty of isolating scales, many biological and social systems show astonishing regularities, West said. For example, for a wide range of species, on average, the "metabolic rate scales with the mass of the organism as a power law of 3/4," he said. "That comes from analyzing the universal properties of networks that support life: the respiratory system, the circulatory system."

West has found that similar laws characterize cities, with their hierarchical networks for transporting people and goods. "It's much more, and to some extent less, than the sum of its parts, and it has all kinds of interesting emergent properties." The prevalence of these scaling laws, West said, supports the idea that these disparate systems are in some sense as efficient as possible. Even in biological evolution, which is not goal-directed, "the continuous feedback mechanisms that are implicit in natural selection leads to something being optimized."

Strevens put it more generally: "the adaptiveness of life is very strong evidence that regularity and simplicity and stability is more the rule than the exception in complex systems." The reason for the stability may even be comprehensible, he speculated. "For an explanation to have such a wide range of applicability in explaining stability, it must be, in some sense a kind of simple explanation."

"You don't need predictive power to have an explanation," Strevens countered. "We understand a lot of things in human history, have a real grasp of why they came out the way they did, but we would never have been able to predict them ahead of the fact."

A general theory?

"The bigger question is: is there a general theory of complexity, and in particular of adaptive systems?" West said. "I'm actually quite skeptical. Even though there have to be principles and laws, maybe to put it into the analog to equations is something that we will never do."

"There's only certain things one brain can comprehend," Chazelle said. Although physics has been very successful in addressing its chosen problems, "in biology, or other fields, it's not obvious to me that that has to be the case." Nonetheless, in view of the success of evolution, he said, "it's hard to believe that there are not some graspable principles that we have not yet gotten our hands on."

Something deserving the name of "law," West said, should make verifiable predictions. But "maybe that's not what biology is about," Gleiser cautioned, because "it depends on historical contingencies. It may be that you just can't do it."

"You don't need predictive power to have an explanation," Strevens countered. "We understand a lot of things in human history, have a real grasp of why they came out the way they did, but we would never have been able to predict them ahead of the fact." The description may even be comprehensible, he speculated. "For an explanation to have such a wide range of applicability in explaining stability, it must be, in some sense a kind of simple explanation."

Extending this framework to the social interactions that really matter to us, "love, and hate, and envy, and excitement, and awe," is a huge challenge, Wests admitted. "How can we ever imagine that we're going to have a scientific theory that explains all of that?" Nonetheless, he added, "it would be a mistake not to try."

Cosmologist Max Tegmark viewed the issue similarly. "Consciousness is the way information feels when it's being processed in a certain very complex way. To me the great mystery is what precisely are these complex ways? What is it that's required?" Tegmark regards feeling as compelling evidence "I have this subjective experience. I know that even more directly than I know that I'm made of quarks. What principle it is that the motion of my quarks satisfy, for there to be that subjective experience?"

Chalmers cautioned that "our best theories of the mind go with our best, latest technologies, so at one point the model of the mind is a steam engine, or is a telephone exchange. These days we're in the information age, so it's very natural to try to understand consciousness in terms of information. But I think there's more to it than that," he said, noting that the informational structure of the world seems to correlate with states of the brain.

No magic in the machine

For the other two panelists, the subjective experience of consciousness, of a seemingly independent "I," is probably a byproduct of oversimplified self-models. "There is no metaphysical gap" that requires explanation, said neuroscientist Michael Graziano. "Our question is not how do you get the inner feeling but how do you get brains that insist they have an inner feeling?"

Source: illustratedprogress.com
Credit: Cornell University

Graziano described a patient of a therapist friend who "thought he had a squirrel in his head instead of a brain." No rational explanations could dissuade him, Graziano said, "because they didn't explain how the squirrel got there." Although most our self-models are not as bad as that, "the way the brain models that or simulates or describes that to itself is a little bit cartoonish, and we describe it as having an inner feeling," Graziano said. "The machine mistakenly thinks it has magic inside of it."

Mechanical engineer Hod Lipson has built robots with varying degrees of self-simulation, which he sees as a working definition of consciousness.

"The notion of feeling is exactly aligned with self-simulation," he said. "It doesn't prove that it's the same thing, but I don't think it's a contradiction."

Lipson emphasizes that self-modeling can vary widely in accuracy. "There's a whole continuum," he said, and even models that are wrong "allow these robots to make the right decision." There's no reason to regard human consciousness is the endpoint of this continuum, he stressed, and focusing on it could be a mistake. Modeling more primitive consciousness is "pretty simple," he said. "We're trying to bite off too much trying to explain human consciousness right away."

Experimental tests

Both Tegmark and Graziano think that their views about consciousness can be experimentally tested. Tegmark has been working with Giulio Tononi of the University of Wisconsin on "integrated information theory," which Tegmark describes as "the first time I actually saw someone write down an equation that information processing was supposed to satisfy to have this magic, subjective quality." With modern neuroscience, he said, "you can now compare what your brain is doing and what you're reporting that you are aware of to test the model's predictions."

These tools are still limited to subjects who can report their subjective experience, however. As Chalmers cautioned, "we don't have direct access to consciousness in any case except for ourselves."

Graziano described consciousness theoretically in terms of a simplified "attention schema," and said the absence of attention leads to specific types of errors in motion control. "This is the kind of test we're doing," he said. "Thus far the theory has failed to be falsified."

Conscious artificial intelligence?

Lipson said artificial-intelligence systems perform better when they include self-modeling—his definition of consciousness. "They are much more resilient. They can act. They can learn faster. They can recover from damage. They can anticipate things before they happen."

But consciousness is irrelevant to whether humans should worry about artificial intelligence, Tegmark said. "You don't really care if that heat-seeking missile coming after you is having an inner experience or not," he quipped. Still, "it matters from an ethical point of view: whether we should treat them also at some level as having ethical rights."

Graziano said consciousness could actually make such systems safer, because it could induce a degree of empathy. "A conscious computer is a computer that could potentially attribute consciousness to us, and therefore it is less likely to go around killing us all."

To frame this issue, researchers use the "Drake equation," named for radio astronomer Dr. Frank Drake, which expresses the probability of contact as a product of many factors. Some factors quantify the planets where life could arise, while others describe the probability that it did so and that it evolved into intelligent, technological creatures who are willing and able to communicate with us.

Recent direct observations, for example from the Kepler space telescope, have clarified the first, physical factors by revealing exoplanets orbiting other stars. "Pretty much every star you see in the sky has at least one planet going around it," said astrophysicist Adam Frank. Moreover, many of these planets are likely to be in the "Goldilocks Zone" where temperatures are not too hot or cold. "The closest planet that's habitable is almost certainly within ten light years," said astrophysicist Jason Thomas Wright.

In view of this results, "it's time to stop snickering about aliens," Frank said. "We can really start asking questions about life (not necessarily intelligent life)." Researchers are beginning to measure chemical makeup of exoplanet atmospheres, and "within 30 years we may have data that tells that there are biospheres," he said.

Astrobiologists like Louisa Preston expect that extraterrestrial life will probably be carbon-based. But rather than a society-changing encounter with a super-advanced civilization, she said, "chances are the first life we find will be a lipid or a bit of protein on Mars. So all the scientists will be out at the bar celebrating," but the rest of the world may be unimpressed.

Intelligent life

What excites non-scientists is the possibility of extraterrestrial intelligence. One challenge is that we might not know it when we see it. "We haven't thought very hard about the ways that potentially very advanced civilizations might be detectible," Wright said. "We should be on the lookout for anomalies." He mentioned intriguing recent Kepler observations, flagged by citizen scientists, that suggest enormous structures passing in front of a star. He referred to one such star as "Tabby's star," after Tabetha Boyajian of Yale University, who showed that the anomalies were real. "We don't have any good ideas for what's going on," Wright said.

The panelists largely agreed that attributing this puzzle to an intelligent civilization demands extraordinary proof. "You've got to exhaust [other possibilities] until your fingers have gone bloody before you would go for the aliens," Frank said. Nonetheless, he thinks that intelligent civilizations are highly likely, based on the Drake equation and the new exoplanet data. Frank and collaborator Woody Sullivan recently calculated that if the odds of life forming on a random planet are "better than one in ten billion trillion," then "we're not the first time that an intelligent civilization has occurred."

But the most important factor in the Drake equation may be the average lifetime of advanced civilization. "If that number is 10,000 years," Frank said. "the galaxy is pretty sterile." To achieve long-term survival, civilizations have to find their way past challenges such as climate change, which he referred to as like our "final exam": "If we make it, we get to go on."

Philosopher Stephen Gardiner agreed that "there are certain kinds of tests that evolving species face," with climate change perhaps being only one. Such "genuinely global, strongly intergenerational" challenges require "a level of sophistication that's not only scientific but also institutional and ethical," and which our response to climate change has not so far exhibited. Gardiner also suggested that extraterrestrials could be waiting for us to show not only technological skills but moral advancement. "They might expect us to reach some ethical threshold before they would permit contact," because contact with us "might be like giving an Uzi to a two-year-old."

Still, even if only a small number of civilizations survive for hundreds of millions of years, they will dominate the average, Wright stressed. "It's really hard for me to imagine that it's inevitable that every species will destroy itself forever." Moreover, once a species has developed settlements on other planets or around other stars, he said, it becomes "pretty immortal."

Leaving Earth

Several panelists commended science-fiction writers for addressing issues such as colonization, but the challenges are daunting. "Interstellar travel is a long, long, long way off," Frank said. "The solar system is home for the next few thousand years."

Even our local neighborhood is extremely inhospitable, and "terraforming" Mars, for example, would likely take thousands of years, Preston noted. Ironically, "to make Mars earthlike we actually need to cause climate change on it," she said, although she also wondered if we have the right to do so. Wright also cautioned that we have yet to establish any settlements in Antarctica, "and it's much more hospitable than any place off the planet."

"Everyone in the community seems to agree that this is not a problem, but they completely disagree about why it's not a problem," Aronson said. The "firewall" paper proposed that space and time cease to exist at the event horizon, which contradicts the longstanding belief that nothing strange happens to spacetime there.

An alternative proposal noted that revealing the entanglement would require an enormous computation. Aronson had previously shown that this type of calculation would be hard even with a quantum computer, so the black hole would have long since evaporated. "If you believe that there are reasonable limits of what a quantum computer can do, then creating this firewall is exponentially hard," he said, "and maybe we should worry about it less than you would have thought."

Soup-can physics

Daniel Harlow co-authored the paper that proposed this calculational escape route, but he spoke on this panel about a different, surprising connection involving quantum information. This work builds on the mapping, identified almost 20 years ago, between certain gravity-free quantum theories, known as conformal field theories (CFTs), and some gravitational models. The correspondence has only been established for universes that have a negative value for the cosmological constant (roughly equivalent to "dark energy"), unlike its apparently positive value in our universe. "This is a theory of somebody's quantum gravity, but it's not our quantum gravity," Harlow said. Nonetheless, theorists expect that many of the results for this "anti-de Sitter" space (AdS) will apply more generally.

Intriguingly, the two connected models differ by one spatial dimension. Harlow used the analogy of a can of soup. "The soup of quantum gravity is actually equal to a theory which just lives on the metal part of the can. But it's a theory without gravity." Harlow stressed that knowing what is happening deep in the "soup" requires an extensive description at the surface.

Surprisingly, the relationship between bulk and surface is similar to that exploited in error-correcting codes that could enable robust quantum computation. "The mathematics people developed to describe how to protect your quantum computation is the same mathematics as this AdS-CFT correspondence," Harlow said. "Maybe the reason a quantum computer will work is because it's simulating quantum gravity." "This is a connection between things that seem like they should have nothing to do with one another," Harlow observed. Moreover, it could produce codes "that have different properties than those coding people have already come up with. For some things they're more efficient."

Experimental tests

AdS/CFT allows researchers to learn about gravitational systems by studying the corresponding quantum systems. But Brian Swingle started out using the correspondence in the other direction. "I was on the boundary," he said, studying quantum electronic systems. In this case, looking deep inside the "soup" gives information about entanglement between "electrons that are farther apart, or larger groups of electrons."

"I was doing quantum gravity without even knowing it." Swingle said. "This was very exciting, because it opened up the possibility that we could actually do experiments to study some of these crazy things that Daniel and Scott have been talking about," Swingle said.

As one example, in modeling the random radiation emitted by black holes, theorists assume that new information is rapidly dispersed. "It's believed that black holes do this as fast as possible; that nothing in nature can scramble information more rapidly," Swingle said. To test this experimentally, researchers simply have to take the corresponding quantum system back in time, perturb it slightly, and then return it to the present to see how extensive the changes are. In this way, "we can understand something about how the black hole scrambles information," he said.

"That may sound kind of crazy," but by using controlled low-temperature systems, "we can actually do things like this." Quantum-optics researchers, for example, routinely apply pulses of light or microwaves to effectively reverse the dynamics of quantum degrees of freedom. "You can, in effect, rewind and fast forward the tape all you want," Swingle said. "The very exciting possibility is that we could actually learn about black holes in a tabletop experiment sometime soon."