When Rebecca Goldin spoke to a recent class of incoming freshmen at George Mason University, she relayed a disheartening statistic: According to a recent study, 36 percent of college students don’t significantly improve in critical thinking during their four-year tenure. “These students had trouble distinguishing fact from opinion, and cause from correlation,” Goldin explained.

Quanta Magazine

About

Original story reprinted with permission from Quanta Magazine, an editorially independent publication of the Simons Foundation whose mission is to enhance public understanding of science by covering research developments and trends in mathematics and the physical and life sciences.

She went on to offer some advice: “Take more math and science than is required. And take it seriously.” Why? Because “I can think of no better tool than quantitative thinking to process the information that is thrown at me.” Take, for example, the study she had cited. A first glance, it might seem to suggest that a third of college graduates are lazy or ignorant, or that higher education is a waste. But if you look closer, Goldin told her bright-eyed audience, you’ll find a different message: “Turns out, this third of students isn’t taking any science.”

Goldin, a professor of mathematical sciences at George Mason, has made it her life’s work to improve quantitative literacy. In addition to her research and teaching duties, she volunteers as a coach at math clubs for elementary- and middle-school students. In 2004, she became the research director of George Mason’s Statistical Assessment Service, which aimed “to correct scientific misunderstanding in the media resulting from bad science, politics or a simple lack of information or knowledge.” The project has since morphed into STATS (run by the nonprofit Sense About Science USA and the American Statistical Association), with Goldin as its director. Its mission has evolved too: It is now less of a media watchdog and focuses more on education. Goldin and her team run statistics workshops for journalists and have advised reporters at publications including FiveThirtyEight, ProPublica and The Wall Street Journal.

When Quanta first reached out to Goldin, she worried that her dual “hats”—those of a mathematician and a public servant—were too “radically different” to reconcile in one interview. In conversation, however, it quickly became apparent that the bridge between these two selves is Goldin’s conviction that mathematical reasoning and study is not only widely useful, but also pleasurable. Her enthusiasm for logic—whether she’s discussing the manipulation of manifolds in high-dimensional spaces or the meaning of statistical significance—is infectious. “I love, love, love what I do,” she said. It’s easy to believe her—and to want some of that delight for oneself.

Quanta Magazine spoke with Goldin about finding beauty in abstract thought, how STATS is arming journalists with statistical savvy, and why mathematical literacy is empowering. An edited and condensed version of the conversation follows.

Where does your passion for mathematics and quantitative thought come from?

As a young person I never thought I liked math. I absolutely loved number sequences and other curious things that, in retrospect, were very mathematical. At the dinner table, my dad, who is a physicist, would pull out some weird puzzle or riddle that sometimes only took a minute to solve, and other times I’d be like, “Huh, I have no idea how that one works!” But there was an overall framework of joy around solving it.

When did you recognize you could apply that excitement about puzzles to pursuing math professionally?

Actually very late in the game. I was always very strong in math, and I did a lot of math in high school. This gave me the false sense that I knew what math was about: I felt like every next step was a little bit more of the same, just more advanced. It was very clear in my mind that I didn’t want to be a mathematician.

But when I went to college at Harvard, I took a course in topology, which is the study of spaces. It wasn’t like anything I’d seen before. It wasn’t calculus; it wasn’t complex calculations. The questions were really complicated and different and interesting in a way I had never expected. And it was just kind of like I fell in love.

You study primarily symplectic and algebraic geometry. How do you describe what you do to people who aren’t mathematicians?

One way I might describe it is to say that I study symmetries of mathematical objects. This comes about when you’re interested in things like our universe, where the Earth is rotating, and it’s also rotating around the sun, and the sun is in a larger system that is rotating. All those rotations are symmetries. There are a lot of other ways symmetries come up, and they can get really, really complicated. So we use neat mathematical objects to think about them, called groups. This is useful because if you’re trying to solve equations, and you know you have symmetries, you can essentially find a way mathematically to get rid of those symmetries and make your equations simpler.

What motivates you to study these complex symmetries?

I just think they’re really beautiful. A lot of mathematics ultimately is artistic rather than useful. Sometimes you see a picture that’s got a lot of symmetry in it, like an M.C. Escher sketch, and it’s like, “Wow, that’s just so amazing!” But when you study mathematics, you start to “see” things in higher dimensions. You’re not necessarily visualizing them in the same way that you could with a sculpture or piece of art. But you start to feel like this whole system of objects that you’re looking at, and the symmetries it has, are really just beautiful. There’s no other good word.

How did you get involved with STATS?

When I arrived as a professor at George Mason, I knew I wanted to do more than research and mathematics. I love teaching, but I felt like I wanted to do something for the world that was not part of the ivory tower of just solving problems that I thought were really curious and interesting.

When I first joined what became STATS, it was a little bit more “gotcha” work: looking at how the media talks about science and mathematics and pointing out when someone has gotten it wrong. As we’ve evolved, I’ve become more and more interested in how journalists think about quantitative issues and how they process them. We found pretty early in our work that there was this huge gap of knowledge and education: Journalists were writing about things that had quantitative content, but they often didn’t absorb what they were writing about, and didn’t understand it, and didn’t have any way to do better because they were often on really tight timelines with limited resources.

So how has your work at STATS changed?

Our mission at STATS has changed to focus on offering journalists two things. One is to be available to answer quantitative questions. They could be as simple as “I don’t know how to calculate this percentage,” or they could be pretty sophisticated things, like “I’ve got this data, and I want to apply this model to it, and I just want to make sure that I’m handling the outliers correctly.” The other really cool thing that we do is, we go to individual news agencies and offer workshops on things like confidence intervals, statistical significance, p values, and all this highly technical language.

Someone once described to me the advice he gives to journalists. He says, “You should always have a statistician in your back pocket.” That’s what we hope to be.

What are the most common pitfalls of reporting on statistics?

A favorite one is distinguishing between causation and correlation. People say, “Oh, that’s obvious. Of course there’s a difference between those two things.” But when you get into examples that target our belief system, it’s really hard to disassociate them. Part of the problem, I think, is that scientists themselves always want to know more than they can with the tools they have. And they don’t always make clear that the questions they’re answering aren’t necessarily the ones you might think they’re answering.

What do you mean?

Like, you might be interested in knowing whether taking hormones is helpful or harmful to women who are postmenopausal. So you start out with a question that’s really well-defined: Does it help or hurt? But you can’t necessarily answer that question. What you can answer is the question of whether women who take hormones whom you enroll in your study — those specific women — have an increase or decrease in, say, heart disease rates or breast cancer rates or stroke rates compared to a control group or to the general population. But that may not answer your initial question, which is: “Is that going to be the case for me? Or people like me? Or the population as a whole?”

What do you hope STATS will achieve?

Partly our goal is to help change the culture of journalism so that people recognize the importance of using quantitative arguments and thinking about quantitative issues before they come to conclusions. That way, they’re coming to conclusions that are supported by science rather than using a study to further their own agenda — which is something scientists do too; they may push a certain interpretation of something. We want to arm journalists with a certain amount of rigor in their thinking so they can challenge a scientist who might say, “Well, you just don’t understand my sophisticated statistic.” There’s a lot of value in giving reporters the tools to develop their sense of quantitative skepticism so that they’re not just bullied.

You argue that statistical literacy gives citizens a kind of power. What do you mean?

What I mean is that if we don’t have the ability to process quantitative information, we can often make decisions that are more based on our beliefs and our fears than based on reality. On an individual level, if we have the ability to think quantitatively, we can make better decisions about our own health, about our own choices with regard to risk, about our own lifestyles. It’s very empowering to not be scared or bullied into doing things one way or another.

On a collective level, the impact of being educated in general is huge. Think about what democracy would be if most of us couldn’t read. We aspire to a literate society because it allows for public engagement, and I think this is also true for quantitative literacy. The more we can get people to understand how to view the world in a quantitative way, the more successful we can be at getting past biases and beliefs and prejudices.

You’ve also said that getting people to understand statistics requires more than reciting numbers. Why do you think storytelling is important for conveying statistical concepts?

As human beings, we live in stories. It doesn’t matter how quantitative you are, we’re all influenced by stories. They become like statistics in our mind. So if you report the statistics without the story, you don’t get nearly the level of interest or emotion or willingness to engage with the ideas.

How has the media’s use of data changed in the 13 years you’ve been with STATS?

With the internet, we see a tremendous growth in data produced by search engines. Journalists are becoming much more adept at collecting these kinds of data and using them in media articles. I think that the current president is also causing a lot of reflection on what we mean by facts, and in that sense journalists maybe think of it as more important in general to get the facts right.

That’s interesting. So you think the public’s awareness of “fake” news and “alternative” facts is motivating journalists to be more rigorous about fact checking?

I do think it’s very motivating. Of course sometimes information gets spun. But ultimately a very small percentage of journalists do that. I think 95 percent of both journalists and scientists are really working hard to get it right.

I’m surprised you’re not more jaded about the media.

Ha! This is maybe more a life view. I think there are people who are pessimistic about humankind and people who are optimistic.

More Quanta

• Kevin Hartnett

  The Mathematician Who Will Make You Fall in Love With Numbers

• Kevin Hartnett

  A Math Genius Blooms Late and Conquers His Field

• Kevin Hartnett

  Mathematicians Are Building a Unified Theory of Geometric Randomness

You also volunteer with math clubs for kids. What ideas about math and math culture do you try to get across?

I try to bring in problems that are really different and fun and curious and weird. For example, I’ve done an activity with kids where I’ve brought in a bunch of ribbons, and I had them learn a little bit about a field called knot theory. There are two things I’m trying to get across to them. One is that math in school is not the whole story—there’s this whole other world that is logical but also beautiful and creative. The second message is a certain emotional framework that I have to offer: that math is a joyous experience.

Original story reprinted with permission from Quanta Magazine, an editorially independent publication of the Simons Foundation whose mission is to enhance public understanding of science by covering research developments and trends in mathematics and the physical and life sciences.

Related Video

Science

The Fascinating Math Behind Why You Won't Win The Powerball

The Powerball jackpot is over a billion dollars but what are your chances?

Suppose aliens land on our planet and want to learn our current scientific knowledge. I would start with the 40-year-old documentary Powers of Ten. Granted, it’s a bit out of date, but this short film, written and directed by the famous designer couple Charles and Ray Eames, captures in less than 10 minutes a comprehensive view of the cosmos.

Quanta Magazine

About

Original story reprinted with permission from Quanta Magazine, an editorially independent publication of the Simons Foundation whose mission is to enhance public understanding of science by covering research developments and trends in mathematics and the physical and life sciences.

The script is simple and elegant. When the film begins, we see a couple picnicking in a Chicago park. Then the camera zooms out. Every 10 seconds the field of vision gains a power of 10—from 10 meters across, to 100, to 1,000 and onward. Slowly the big picture reveals itself to us. We see the city, the continent, Earth, the solar system, neighboring stars, the Milky Way, all the way to the largest structures of the universe. Then in the second half of the film, the camera zooms in and delves into the smallest structures, uncovering more and more microscopic details. We travel into a human hand and discover cells, the double helix of the DNA molecule, atoms, nuclei and finally the elementary quarks vibrating inside a proton.

The movie captures the astonishing beauty of the macrocosm and microcosm, and it provides the perfect cliffhanger endings for conveying the challenges of fundamental science. As our then-8-year-old son asked when he first saw it, “How does it continue?” Exactly! Comprehending the next sequence is the aim of scientists who are pushing the frontiers of our understanding of the largest and smallest structures of the universe. Finally, I could explain what Daddy does at work!

Powers of Ten also teaches us that, while we traverse the various scales of length, time and energy, we also travel through different realms of knowledge. Psychology studies human behavior, evolutionary biology examines ecosystems, astrophysics investigates planets and stars, and cosmology concentrates on the universe as a whole. Similarly, moving inward, we navigate the subjects of biology, biochemistry, and atomic, nuclear and particle physics. It is as if the scientific disciplines are formed in strata, like the geological layers on display in the Grand Canyon.

Moving from one layer to another, we see examples of emergence and reductionism, these two overarching organizing principles of modern science. Zooming out, we see new patterns “emerge” from the complex behavior of individual building blocks. Biochemical reactions give rise to sentient beings. Individual organisms gather into ecosystems. Hundreds of billions of stars come together to make majestic swirls of galaxies.

As we reverse and take a microscopic view, we see reductionism at work. Complicated patterns dissolve into underlying simple bits. Life reduces to the reactions among DNA, RNA, proteins and other organic molecules. The complexity of chemistry flattens into the elegant beauty of the quantum mechanical atom. And, finally, the Standard Model of particle physics captures all known components of matter and radiation in just four forces and 17 elementary particles.

Which of these two scientific principles, reductionism or emergence, is more powerful? Traditional particle physicists would argue for reductionism; condensed-matter physicists, who study complex materials, for emergence. As articulated by the Nobel laureate (and particle physicist) David Gross: Where in nature do you find beauty, and where do you find garbage?

Take a look at the complexity of reality around us. Traditionally, particle physicists explain nature using a handful of particles and their interactions. But condensed matter physicists ask: What about an everyday glass of water? Describing its surface ripples in terms of the motions of the roughly 10²⁴ individual water molecules—let alone their elementary particles—would be foolish. Instead of the impenetrable complexities at small scales (the “garbage”) faced by traditional particle physicists, condensed matter physicists use the emergent laws, the “beauty” of hydrodynamics and thermodynamics. In fact, when we take the number of molecules to infinity (the equivalent of maximal garbage from a reductionist point of view), these laws of nature become crisp mathematical statements.

While many scientists praise the phenomenally successful reductionist approach of the past centuries, John Wheeler, the influential Princeton University physicist whose work touched on topics from nuclear physics to black holes, expressed an interesting alternative. “Every law of physics, pushed to the extreme, will be found to be statistical and approximate, not mathematically perfect and precise,” he said. Wheeler pointed out an important feature of emergent laws: Their approximate nature allows for a certain flexibility that can accommodate future evolution.

In many ways, thermodynamics is the gold standard of an emergent law, describing the collective behavior of a large number of particles, irrespective of many microscopic details. It captures an astonishingly wide class of phenomena in succinct mathematical formulas. The laws hold in great universality—indeed, they were discovered before the atomic basis of matter was even established. And there are no loopholes. For example, the second law of thermodynamics states that a system’s entropy—a measure of the amount of hidden microscopic information—will always grow in time.

Modern physics provides a precise language to capture the way things scale: the so-called renormalization group. This mathematical formalism allows us to go systematically from the small to the large. The essential step is taking averages. For example, instead of looking at the behavior of individual atoms that make up matter, we can take little cubes, say 10 atoms wide on each side, and take these cubes as our new building blocks. One can then repeat this averaging procedure. It is as if for each physical system one makes an individual Powers of Ten movie.

Renormalization theory describes in detail how the properties of a physical system change if one increases the length scale on which the observations are made. A famous example is the electric charge of particles that can increase or decrease depending on quantum interactions. A sociological example is understanding the behavior of groups of various sizes starting from individual behavior. Is there wisdom in crowds, or do the masses behave less responsibly?

Most interesting are the two endpoints of the renormalization process: the infinite large and infinite small. Here things will typically simplify because either all details are washed away, or the environment disappears. We see something like this with the two cliffhanger endings in Powers of Ten. Both the largest and the smallest structures of the universe are astonishingly simple. It is here that we find the two “standard models,” of particle physics and cosmology.

Remarkably, modern insights about the most formidable challenge in theoretical physics—the push to develop a quantum theory of gravity—employ both the reductionist and emergent perspectives. Traditional approaches to quantum gravity, such as perturbative string theory, try to find a fully consistent microscopic description of all particles and forces. Such a “final theory” necessarily includes a theory of gravitons, the elementary particles of the gravitational field. For example, in string theory, the graviton is formed from a string that vibrates in a particular way. One of the early successes of string theory was a scheme to compute the behavior of such gravitons.

However, this is only a partial answer. Einstein taught us that gravity has a much wider scope: It addresses the structure of space and time. In a quantum-mechanical description, space and time would lose their meaning at ultrashort distances and time scales, raising the question of what replaces those fundamental concepts.

A complementary approach to combining gravity and quantum theory started with the groundbreaking ideas of Jacob Bekenstein and Stephen Hawking on the information content of black holes in the 1970s, and came into being with the seminal work of Juan Maldacena in the late 1990s. In this formulation, quantum space-time, including all the particles and forces in it, emerges from a completely different “holographic” description. The holographic system is quantum mechanical, but doesn’t have any explicit form of gravity in it. Furthermore, it typically has fewer spatial dimensions. The system is, however, governed by a number that measures how large the system is. If one increases that number, the approximation to a classical gravitational system becomes more precise. In the end, space and time, together with Einstein’s equations of general relativity, emerge out of the holographic system. The process is akin to the way that the laws of thermodynamics emerge out of the motions of individual molecules.

More Quanta

• Natalie Wolchover

  Mathematicians Bridge the Divide Between Infinity and the Physical World

• Natalie Wolchover

  The Man Who's Trying to Kill Dark Matter

• Natalie Wolchover

  Quantum Gravity Research Could Unearth the True Nature of Time

In some sense, this exercise is exactly the opposite of what Einstein tried to achieve. His aim was to build all of the laws of nature out of the dynamics of space and time, reducing physics to pure geometry. For him, space-time was the natural “ground level” in the infinite hierarchy of scientific objects—the bottom of the Grand Canyon. The present point of view thinks of space-time not as a starting point, but as an end point, as a natural structure that emerges out of the complexity of quantum information, much like the thermodynamics that rules our glass of water. Perhaps, in retrospect, it was not an accident that the two physical laws that Einstein liked best, thermodynamics and general relativity, have a common origin as emergent phenomena.

In some ways, this surprising marriage of emergence and reductionism allows one to enjoy the best of both worlds. For physicists, beauty is found at both ends of the spectrum.

Animals are living color. Wasps buzz with painted warnings. Birds shimmer their iridescent desires. Fish hide from predators with body colors that dapple like light across a rippling pond. And all this color on all these creatures happened because other creatures could see it.

The natural world is so showy, it’s no wonder scientists have been fascinated with animal color for centuries. Even today, the questions how animals see, create, and use color are among the most compelling in biology.

Until the last few years, they were also at least partially unanswerable—because color researchers are only human, which means they can’t see the rich, vivid colors that other animals do. But now new technologies, like portable hyperspectral scanners and cameras small enough to fit on a bird’s head, are helping biologists see the unseen. And as described in a new Science paper, it's a whole new world.

Visions of Life

The basics: Photons strike a surface—a rock, a plant, another animal—and that surface absorbs some photons, reflects others, refracts still others, all according to the molecular arrangement of pigments and structures. Some of those photons find their way into an animal’s eye, where specialized cells transmit the signals of those photons to the animal’s brain, which decodes them as colors and shapes.

It's the brain that determines whether the colorful thing is a distinct and interesting form, different from the photons from the trees, sand, sky, lake, and so on it received at the same time. If it’s successful, it has to decide whether this colorful thing is food, a potential mate, or maybe a predator. “The biology of color is all about these complex cascades of events,” says Richard Prum, an ornithologist at Yale University and co-author of the paper.

In the beginning, there was light and there was dark. That is, basic greyscale vision most likely evolved first, because animals that could anticipate the dawn or skitter away from a shadow are animals that live to breed. And the first eye-like structures—flat patches of photosensitive cells—probably didn't resolve much more than that. It wasn't enough. “The problem with using just light and dark is that the information is quite noisy, and one problem that comes up is determining where one object stops and another one starts. ” says Innes Cuthill, a behavioral ecologist at the University of Bristol and coauthor of the new review.

Color adds context. And context on a scene is an evolutionary advantage. So, just like with smart phones, better resolution and brighter colors became competitive enterprises. For the resolution bit, the patch light-sensing cells evolved over millions of years into a proper eye—first by recessing into a cup, then a cavity, and eventually a fluid-filled spheroid capped with a lens. For color, look deeper at those light-sensing cells. Wedged into their surfaces are proteins called opsins. Every time they get hit with a photon—a quantum piece of light itself—they transduce that signal into an electrical zap to the rudimentary animal's rudimentary brain. The original light/dark opsin mutated into spin-offs that could detect specific ranges of wavelengths. Color vision was so important that it evolved independently multiple times in the animal kingdom—in mollusks, arthropods, and vertebrates.

In fact, primitive fish had four different opsins, to sense four spectra—red, green, blue, and ultraviolet light. That four-fold ability is called tetrachromacy, and the dinosaurs probably had it. Since they're the ancestors of today’s birds, many of them are tetrachromats, too.

But modern mammals don't see things that way. That's probably because early mammals were small, nocturnal things that spent their first 100 million years running around in the dark, trying to keep from being eaten by tetrachromatic dinosaurs. “During that period the complicated visual system they inherited from their ancestors degraded,” says Prum. “We have a clumsy, retrofitted version of color vision. Fishes, and birds, and many lizards see a much richer world than we do."

In fact, most monkeys and apes are dichromats, and see the world as greyish and slightly red-hued. Scientists believe that early primates regained three-color vision because spotting fresh fruit and immature leaves led to a more nutritious diet. But no matter how much you enjoy springtime of fall colors, the wildly varicolored world we humans live in now isn't putting on a show for us. It's mostly for bugs and birds. “Flowering plants of course have evolved to signal pollinators,” says Prum. “The fact that we find them beautiful is incidental, and the fact that we can see them at all is because of an overlap in the spectrums insects and birds can see and the ones we can see.”

Covered in Color

And as animals gained the ability to sense color, evolution kickstarted an arms race in displays—hues and patterns that aided in survival became signifiers of ace baby-making skills. Almost every expression of color in the natural world came about to signal, or obscure, a creature to something else.

For instance, "aposematism" is color used as a warning—the butterfly’s bright colors say “don’t eat me, you'll get sick.” "Crypsis" is color used as camouflage. Color serves social purposes, too. Like, in mating. Did you know that female lions prefer brunets? Or that paper wasps can recognize each others’ faces? “Some wasps even have little black spots that act like karate belts, telling other wasps not to try and fight them,” says Elizabeth Tibbetts, an entomologist at the University of Michigan.

Related Stories

• Adam Rogers

  The Science of Why No One Agrees on the Color of This Dress

• Neel Patel

  3-D Map Shows the Colors You See But Can't Name

• Neel Patel

  No, These Biohackers Can't Give Themselves Infrared Vision

But animals display colors using two very different methods. The first is with pigments, colored substances created by cells called chromatophores (in reptiles, fish, and cephalopods), and melanocytes (in mammals and birds). They absorb most wavelengths of light and reflect just a few, limiting both their range and brilliance. For instance, most animals cannot naturally produce red; they synthesize it from plant chemicals called carotenoids.

The other way animals make color is with nanoscale structures. Insects, and, to a lesser degree, birds, are the masters of color-based structure. And compared to pigment, structure is fabulous. Structural coloration scatters light into vibrant, shimmering colors, like the shimmering iridescent bib on a Broad-tailed hummingbird, or the metallic carapace of a Golden scarab beetle. And scientists aren't quite sure why iridescence evolved. Probably to signal mates, but still: Why?

Decoding the rainbow of life

The question of iridescence is similar to most questions scientists have about animal coloration. They understand what the colors do in broad strokes, but there's till a lot of nuance to tease out. This is mostly because, until recently, they were limited to seeing the natural world through human eyes. “If you ask the question, what’s this color for, you should approach it the way animals see those colors,” says Tim Caro, a wildlife biologist at UC Davis and the organizing force behind the new paper. (Speaking of mysteries, Caro recently figured out why zebras have stripes.)

Take the peacock. “The male’s tail is beautiful, and it evolved to impress the female. But the female may be impressed in a different way than you or I,” Caro says. Humans tend to gaze at the shimmering eyes at the tip of each tail feather; peahens typically look at the base of the feathers, where they attach to the peacock’s rump. Why does the peahen find the base of the feathers sexy? No one knows. But until scientists strapped to the birds' heads tiny cameras spun off from the mobile phone industry, they couldn't even track the peahens' gaze.

Another new tech: Advanced nanomaterials give scientists the ability to recreate the structures animals use to bend light into iridescent displays. By recreating those structures, scientists can figure out how genetically expensive they are to make.

Likewise, new magnification techniques have allowed scientists to look into an animal’s eye structure. You might have read about how mantis shrimp have not three or four but a whopping 12 different color receptors, and how they see the world in psychedelic hyperspectral saturation. This isn’t quite true. Those color channels aren’t linked together—not like they are in other animals. The shrimp probably aren’t seeing 12 different, overlapping color spectra. “We are thinking maybe those color receptors are being turned on or off by some other, non-color, signal,” says Caro.

But perhaps the most important modern innovation in biological color research is getting all the different people from different disciplines together. “There are a lot of different sorts of people working on color,” says Caro. “Some behavioral biologists, some neurophysiologists, some anthropologists, some structural biologists, and so on.”

And these scientists are scattered all over the globe. He says the reason he brought everyone to Berlin is so they could finally synthesize all these sub-disciplines together, and move into a broader understanding of color in the world. The most important technology in understanding animal color vision isn't a camera or a nanotech surface. It's an airplane. Or the internet.

Related Video

Science

How the Morpho Butterfly Can Be Blue But Also Not Really Blue

The morpho butterfly appears blue but it isn't actually. It looks blue not because of pigment but because of some very fancy scales on its wings.

Nick Goldschmidt has been lucky so far. A wildfire has burned more than 8,000 acres just north of his vineyards in Geyserville, California, but so far his vines are OK. So is his house in Healdsburg, roughly midway between Geyserville and a 36,000-acre fire that destroyed more than 2,800 homes in Santa Rosa.

But now, amid the charred, empty spaces that scar northern California’s winegrowing region, under skies yellowed by smoke, Goldschmidt has a race to win. Wildfires can ruin the flavor of wine grapes, a problem called smoke taint. “I’ve worked with smoke before,” Goldschmidt says. “It is not an easy thing to fix. But in my experience, it’s more about contact time. So the key thing is, if you have vineyards near the fire, you’ve got to get the grapes off.”

Depending on wind, smoke from the Atlas Fire could potentially reach Goldschmidt’s Napa vineyard, where about 15 percent of the fruit remains to be harvested. He now plans to harvest the rest by the weekend.

That’s typical of Napa, where 80 to 85 percent of the 2017 harvest is done. In nearby Sonoma, 90 percent of the grapes are in. But that still means that a few grapes could get exposed to smoke, and fire and heat could damage the vines. In a region key to California’s $34 billion wine industry—and that figure doesn’t even include the enormous tourist business—that’s a big deal. Fires have killed 31 people so far, destroyed thousands of homes, and consumed the efforts of more than 8,000 firefighters. And the winemakers of the area are trying to make sure the damage to their livelihoods doesn’t get worse.

Winemaking regions around the world, especially in Australia, have been dealing with the consequences of more active fire seasons near vineyards since at least the turn of the century—but the problems haven’t really hit California yet. The state’s frequent fires haven’t intersected with its vineyards. Until now.

Smoke is complicated stuff. Everyone in the Bay Area has gotten a taste in the past few days—that medicinal, ashy, burnt flavor comes from, among other ingredients, molecules called polycyclic aromatic hydrocarbons, nitrogen and sulfur oxides, other organic compounds, and even tiny particles carried aloft by heat and air currents. If you’ve ever sat near a campfire or cooked on a grill, you know it’s not necessarily an unpleasant aroma, as cognitively dissonant as that may feel when you realize it comes from blazes that have destroyed the lives of thousands.

But what’s delicious in bacon or lox generally isn’t—depending on how much you have—in wine. The actual flavor compounds are molecules called volatile phenols. “Volatile” means they evaporate, and in chemistry “phenols” are benzene rings (a hexagon of carbon atoms with hydrogen atoms sticking off like a snowflake) connected to a hydrogen-and-an-oxygen. You might know them better as the aroma of peat in some whiskies or of antiseptic or Band-Aids. Their volatility means that in your mouth, they turn into a vapor that gets sucked up retronasally, through the back of the throat to the sensitive layer of nerve endings behind your nose that translates chemicals into odors.

Smoke taint in grapes has two specific markers: guaiacol and 4-methylguaiacol. They taste like, well, smoke. Expose grapes on the vine to them, and the wine will taste smoky. Obvious, right? Except no. “The mechanism is a little bit unclear,” says Kerry Wilkinson, an oenologist who studies smoke taint at the University of Adelaide. Leaves have pores called stomata involved in respiration, “but when grapevines are exposed to smoke, the stomata close almost immediately and photosynthesis stops,” she says. “The guaiacol conjugates are getting to not only the skins but the pulp of the fruit. I think it’s just permeation, but I don’t think anyone’s done the research.”

Making things even more complicated, grapevines have their own way of dealing with a barbecue. “Those compounds, once they’re taken up, the grapevine will stick one or more sugar molecules on them,” Wilkinson says. “We think that’s to make them less toxic to the plant.” This process—it’s called glycosylation, and the sugars are called glycosides—turns the volatile phenols non-volatile. Which means you can’t taste them in the grape juice.

But ferment that juice into wine, and acids in it will break those sugars off. Poof: Smoke gets in your wines.

I don’t mean to be flip here; smoke taint from the Canberra bushfires of 2003 cost Australian vineyards more than $4 million; fires in 2004 cost another $7 million. Once grapes are tainted, the wine isn’t easy to fix. Those ashy flavors are too strong; you can’t just try to blend in other, untainted wine to cover it up. Efforts to filter it with activated charcoal and reverse osmosis can filter out flavors you might want in the wine, too. Heck, guaiacol and 4-methylguaiacol are markers for oak-barrel aging, too. Nobody ever describes an over-oaked Chardonnay as “smoke-tainted,” but—well, maybe they should, actually. And some grape varietals—shiraz, particularly—already have naturally high guaiacol levels.

All of which might be fine. Most of the grapes were picked before the fires came. In general, “if the fruit’s already been harvested this year, it should be OK,” Wilkinson says. A couple dozen wineries suffered damage so far, from minor to total. But Northern California has almost 250,000 acres of wine grape vines—more than 100,000 of them in Napa and Sonoma Counties as of 2016.

It looks like the grapes those vines produce will be OK next year. “There’s no carryover effect from one season to the other,” Wilkinson says. “We haven't seen any evidence to suggest that any of those smoke compounds are bound up from one season to another in the grapevine.” They get into the grapes, which come off, and the leaves, which fall off or get pruned. (And a little more luck: The grapes left to harvest in Napa are mostly Cabernet Sauvignon, which turns out to be more resistant to smoke taint than some other varietals.)

But vines themselves are sensitive to heat. “They can be scorched, and if it’s severe, that can permanently damage or kill grapevines,” Wilkinson says. “If there’s just a little bit of scorching, vines can recover, but the yield can be decreased in the season immediately after.”

It turns out it’s pretty hard to burn down a California vineyard. In part that’s because most of them are irrigated, so they’re wet and thus resistant to fire. Even when the cover crop growing between the vines gets burned in a fast-moving wildfire, “it’s pretty damn hard” to get a California vineyard to catch, Goldschmidt says.

When he was working in Chile, though, he saw a wind-driven wildfire very much like the ones affecting California destroy a vineyard. “That was devastating,” Goldschmidt says. “A lot of those vineyards are dry-farmed, so they burned much more easily.”

Heat damage is a lot like frost damage, something California vintners know a lot about. Proper pruning and treatment can save an injured vine. The trick is knowing if they’re injured, and how badly. Sometimes vintners will have to cut through the trunks of the vines to assess whether the phloem, the living and respiring part of the wood, is still healthy. But that’s a destructive test—sometimes destroying the vine in an attempt to save it.

So grape-growers look for other ways to assess their vines. “We look at things like, are the irrigation lines melted? Is there indication of scorching of the trunk and canopy? How much fire damage was there to anything growing in between the vines?” Wilkinson says.

That’s an assessment that’ll probably have to wait until the fires are under control. Maybe Goldschmidt’s luck will hold out. “This is my 29th vintage in Sonoma. It’s the first time the Alexander Valley was earlier in maturity than the Napa Valley. Usually it’s 10 days later,” he says. “If it had been the other way I would have really been hammered.”

Ordinarily I might end the story with that, but in this case it’s not as lucky as it sounds. Napa and Sonoma did indeed have a weird year. It rained hard after years of drought, and then over the summer it got really hot. Vintners irrigated when they might not have, which lowered the sugar levels in the grapes as they took up the water…and then it got hot again. “It’s been really hard to make a harvest decision based on sugar,” Goldschmidt says. “It’s been more about flavor and tannin.”

Based on those organoleptic assessments—the fanciest possible way of saying “how it tastes”—most of Napa and Sonoma brought in their fruit in July and August instead of, well, now.

To whom should the vintners send a thank-you? “Over the last five years or so we’ve had this period of very high temperatures that coincided with low precipitation, punctuated by very wet conditions,” says Noah Diffenbaugh, a climate researcher at Stanford. It’s exactly what Diffenbaugh’s group warned would happen in a prescient 2006 paper in the Proceedings of the National Academy of Sciences titled “Extreme heat reduces and shifts United States premium wine production in the 21st century.”

Wildfires

• Joe Eaton

  How Bizarre Is This Year’s Wildfire Season, Really?

• Adam Rogers

  The Napa Fire Is a Perfectly Normal Apocalypse

• Laura Mallonee

  Photo of the Week: Hell Descends on California's Wine Country

Their point? At first, heat and rare-but-extreme rain is going to change how winegrowing regions work. Eventually more northerly regions will be better for grapes—hello, Oregon’s Willamette Valley—and existing grape-growing regions will change the varieties they grow.

It’s in the nature of global warming that extreme climate events will become less rare. “We’ve done a lot of work trying to understand how global warming impacts temperature extremes,” Diffenbaugh says. “The extremes are really where we feel the climate.”

The vagaries of climate change let the 2017 vintage mostly dodge the economic devastation that smoke taint would have caused. It’s a faintly silver lining to the clouds of ash and smoke now parked over thousands of acres of death and destruction. But that silver lining won’t last. These won’t be the last fires; next time, maybe the harvest won’t happen first. The most frightening truth about the extreme climate event that is the northern California fires is that such events won’t always be extreme. They’ll be normal.

Related Video

Culture

Booze Science | Ice

Booze Science is better drinking through chemistry. WIRED articles editor Adam Rogers explores the scientific ways ice can influence a cocktail with Jennifer Colliau, beverage director at San Francisco's innovative bar The Interval at The Long Now.

Inside a red-bricked building on the north side of Washington DC, internist Shantanu Nundy rushes from one examining room to the next, trying to see all 30 patients on his schedule. Most days, five of them will need to follow up with some kind of specialist. And odds are they never will. Year-long waits, hundred-mile drives, and huge out of pocket costs mean 90 percent of America’s most needy citizens can’t follow through on a specialist referral from their primary care doc.

But Nundy’s patients are different. They have access to something most people don’t: a digital braintrust of more than 6,000 doctors, with expert insights neatly collected, curated, and delivered back to Nundy through an artificial intelligence platform. The online system, known as the Human Diagnosis Project, allows primary care doctors to plug into a collective medical superintelligence, helping them order tests or prescribe medications they’d otherwise have to outsource. Which means most of the time, Nundy’s patients wait days, not months, to get answers and get on with their lives.

In the not-too-distant future, that could be the standard of care for all 30 million people currently uninsured or on Medicaid. On Thursday, Human Dx announced a partnership with seven of the country’s top medical institutions to scale up the project, aiming to recruit 100,000 specialists—and their expert assessments—in the next five years. Their goal: close the specialty care gap for 3 million Americans by 2022.

In January, a single mom in her 30s came to see Nundy about pain and joint stiffness in her hands. It had gotten so bad that she had to stop working as a housekeeper, and she was growing desperate. When Nundy pulled up her chart, he realized she had seen another doctor at his clinic a few months prior who referred her to a specialist. But once the patient realized she’d have to pay a few hundred dollars out of pocket for the visit, she didn’t go. Instead, she tried get on a wait list at the public hospital, where she couldn’t navigate the paperwork—English wasn’t her first language.

Now, back where she started, Nundy examined the patient’s hands, which were angrily inflamed. He thought it was probably rheumatoid arthritis, but because the standard treatment can be pretty toxic, he was hesitant to prescribe drugs on his own. So he opened up the Human Dx portal and created a new case description: “35F with pain and joint stiffness in L/R hands x 6 months, suspected AR.” Then he uploaded a picture of her hands and sent out the query.

Within a few hours a few rheumatologists had weighed in, and by the next day they’d confirmed his diagnosis. They’d even suggested a few follow-up tests just to be sure and advice about a course of treatment. “I wouldn’t have had the expertise or confidence to be able to do that on my own,” he says.

Nundy joined Human Dx in 2015, after founder Jayanth Komarneni recruited him to pilot the platform’s core technologies. But the goal was always to go big. Komarneni likens the network to Wikipedia and Linux, but instead of contributors donating encyclopedia entries or code, they donate medical expertise. When a primary care doc gets a perplexing patient, they describe their background, medical history, and presenting symptoms—maybe adding an image of an X-ray, a photo of a rash, or an audio recording of lung sounds. Human Dx’s natural language processing algorithms will mine each case entry for keywords to funnel it to specialists who can create a list of likely diagnoses and recommend treatment.

Now, getting back 10 or 20 different doctors’ takes on a single patient is about as useful as having 20 friends respond individually via email to a potluck invitation. So Human Dx’s machine learning algorithms comb through all the responses to check them against all the project’s previously stored case reports. The network uses them to validate each specialist's finding, weight each one according to confidence level, and combine it with others into a single suggested diagnosis. And with every solved case, Human Dx gets a little bit smarter. “With other online tools if you help one patient you help one patient,” Komarneni says. “What’s different here is that the insights gained for one patient can help so many others. Instead of using AI to replace jobs or make things cheaper we’re using it to provide capacity where none exists.”

Komarneni estimates that those electronic consults can handle 35 to 40 percent of specialist visits, leaving more time for people who really need to get into the office. That’s based on other models implemented around the country at places such as San Francisco General Hospital, UCLA Health System, and Brigham and Women’s Hospital. SFGH’s eReferral system cut the average waiting time for an initial consult from 112 days to 49 within its first year.

That system, which is now the default for every SFGH specialty, relies on dedicated reviewers who get paid to respond to cases in a timely way. But Human Dx doesn’t have those financial incentives—its service is free. Today, though, by partnering with the American Board of Medical Specialities, Human Dx can now offer continuing education and improvement credits to satisfy at least some of the 200 hours doctors are required to complete every four years. And the American Medical Association, the nation’s largest physician group, has committed to getting its members to volunteer, as well as supporting program integrity by verifying physicians on the platform.

Related Stories

• Nick Stockton

  Veritas Genetics Scoops Up an AI Company to Sort Out Its DNA

• Megan Molteni

  Thanks to AI, Computers Can Now See Your Health Problems

• Megan Molteni

  The Chatbot Therapist Will See You Now

It’s a big deal to have the AMA on board. Physicians have historically been wary of attempts to supplant or complement their jobs with AI-enabled tools. But it’s important to not mistake the organization’s participation in the alliance for a formal pro-artificial intelligence stance. The AMA doesn’t yet have an official AI policy, and it doesn’t endorse any specific companies, products, or technologies, including Human Dx’s proprietary algorithms. The medical AI field is still young, with plenty of potential for unintended consequences.

Like discrepancies in quality of care. Alice Chen, the chief medical officer for the San Francisco Health Network and co-director of SFGH’s Center for Innovation in Access and Quality, worries that something like Human Dx might create a two-tiered medical system, where some people get to actually see specialists and some people just get a computerized composite of specialist opinions. “This is the edge of medicine right now,” Chen says. “You just have to find the sweet spot where you can leverage expertise and experience beyond traditional channels and at the same time ensure quality care.”

Researchers at Johns Hopkins, Harvard, and UCSF have been assessing the platform for accuracy and recently submitted results for peer review. The next big hurdle is money. The project is currently one of eight organizations in contention for a $100 million John D. and Catherine T. MacArthur Foundation grant. If Human Dx wins, they’ll spend the money to roll out nationwide. The alliance isn’t contingent on the $100 million award, but it would certainly be a nice way to kickstart the process—especially with specialty visits accounting for more than half of all trips to the doctor’s office.

So it’s possible that the next time you go in for something that stumps your regular physician, instead of seeing a specialist across town, you’ll see five or 10 from around the country. All it takes is a few minutes over lunch or in an elevator to put on a Sherlock Holmes hat, hop into the cloud, and sleuth through your case.

Related Video

Technology

The Robot Will See You Now – AI and Health Care

Artificial intelligence is now detecting cancer and robots are doing nursing tasks. But are there risks to handing over elements of our health to machines, no matter how sophisticated?