MIT Press

Excerpted from Democratizing Our Data: A Manifesto by Julia Lane. Reprinted with permission from The MIT PRESS. Copyright 2020. On sale as an ebook now. On sale in print 9/1/2020.

Nowadays when people have an appointment to go to across town, their calendar app obligingly predicts how long it’s going to take to get there. When they go to Amazon to research books that might be of interest, Amazon makes helpful suggestions—and asks for feedback on how to make its platform better. If they select photos from Google Photos, it suggests people to send them to, prompts with other photos it thinks are like the ones selected, and warns if the zip file is going to be especially big. Our apps today are aware of multiple dimensions of the data they manage for us, they update that information in real time, and suggest options and possibilities based upon those dimensions. In other words, the private sector sets itself up for success because it uses data to provide us with useful products and services.

The government—not so much. Lack of data makes Joe Salvo’s job much more difficult. He is New York City’s chief demographer, and he uses the Census Bureau’s American Community Survey (ACS) data to prepare for emergencies like Hurricane Sandy. He needs to use data to decide how to get older residents to physically accessible shelters—operationally, where to tell a fleet of fifty buses to go to pick up and evacuate seniors. He needs data on the characteristics of the local population for the Mayor’s Office for People with Disabilities. He needs to identify areas with large senior populations to tell the Metropolitan Transit Authority where to send buses. He needs to identify neighborhoods with significant vulnerable populations so that the Department of Health and Mental Hygiene can install emergency generators at Department of Health facilities. But the products produced by the federal statistical system do not provide him with the value that he needs. The most current data from the prime source about the US population, the ACS, is released two years after collection, and that itself reflects five-year moving averages.

Creating value for the consumer is key to success in the private sector. The challenge to statistical agencies is figuring out how to get set up for success and produce high-quality data as measured against the same checklist by providing access to data while at the same time protecting privacy and confidentiality.

The problem is that the checklist for agencies is even longer with additional requirements so that Joe Salvo and his counterparts can do their jobs better. One requirement, given that the United States is a democracy, is that statistics should be as unbiased as possible—so that all residents, whatever their characteristics, are counted and that they are treated equally in measurement. Correcting for the inevitable bias in source data is an important role for statistical agencies. Another requirement is that collecting the data is cost-effective, so that the taxpayer gets a good deal. A third requirement is that the information collected is consistent over time so that trends can easily be spotted and responded to. Agencies need outside help from both stakeholders and experts to ensure all these requirements are met. That requires access to data, which requires dealing with confidentiality issues.

The value that is generated when governmental agencies can straightforwardly provide access and produce new measures can be great. For example, the same people who bring you the National Weather service and its weather predictions—the National Oceanic and Atmospheric Agency, or NOAA—have provided scientists and entrepreneurs with access to data to develop new products, such as predicting forest fires and providing real-time intelligence services for natural disasters in the United States and Canada. City transit agencies share transit data with private-sector app developers who produce high-quality apps that offer real-time maps of bus locations and expected arrival times at bus stops and more.

But other cases, when the government has confidential data, which is the case for most statistical agencies, are different. We need to be able to rely on our government to keep some data very private, but that will often mean that we have to give up on the granularity of government data that are produced. If, for example, the IRS provided so much information about taxpayers that it was possible to know how much money a given individual made, the public would be outraged.

So many government agencies have to worry about two things: (1) producing data that have value and (2) at the same time ensuring that the confidentiality of data owners is protected. This can be done. Some—smaller—governments have succeeded better than others in creating data systems that live up to the checklist of the desired features while at the same time protecting privacy.

Take the child services system as an example. To put child services in context, almost four in ten US children will be referred to their local government for possible child abuse or neglect by the time they’re eighteen. That’s almost four million referrals a year. Frontline caseworkers have to make quick decisions on these referrals. If they are wrong in either direction, the potential downside is enormous: Children incorrectly screened because of inadequate or inaccurate data could be ripped away from loving families. Or, conversely, also as a result of poor data, children could be left with abusive families and die. Furthermore, there could be bias in decisions, leaving black or LGBTQ parents more likely to be penalized, for example.

In 2014, Allegheny County’s Office of Children, Youth and Families (CYF) in Pennsylvania stepped up to the plate to use its internal data in a careful and ethical manner to help caseworkers do their job better. The results have captured national attention, as reported in a New York Times Magazine article. CYF brought in academic experts to design an automatic risk-scoring tool that summarizes information about a family to help the caseworker make better decisions. The risk score, a number between 1 and 20, makes use of a great deal of the information about the family in the county’s system, such as child welfare records, jail records, and behavioral health records, to predict adverse events that can lead to placing a child in foster care.

An analysis of the effectiveness of that tool showed that a child whose placement score at referral is the highest possible—20—is twenty-one times more likely to be admitted to a hospital for a self-inflicted injury, seventeen times more likely to be admitted for being physically assaulted, and 1.4 times more likely to be admitted for suffering from an accidental fall than a child with a risk score of 1, the lowest  possible. An independent evaluation found that caseworker decisions that were informed by the score were more accurate (cases were more likely to be correctly identified as needing help and less likely to be incorrectly identified as not needing help), case workloads decreased, and racial bias was likely to be reduced.  On the eight-item checklist Allegheny County hit on all items. They produced a new product that was used, was cost effective, and produced real-time, accurate, complete, relevant, accessible, interpretable, granular, and consistent data. And CYF didn’t breach confidentiality. But most importantly, Allegheny County worked carefully and openly with advocates for parents, children, and civil rights to ensure that the program was not built behind closed doors. They worked, in other words, to ensure that the new measures were democratically developed and used.

The Allegheny County story is one illustration of how new technologies can be used to democratize the decision of how to balance the ever-present tradeoff between the utility of new measurement against the risk of compromising confidentiality. They took advantage of the potential to create useful information that people and policy makers need while at the same time protecting privacy. That potential can be made real in other contexts by making the value of data clearer to the public. While that utility/cost tradeoff has typically been made by a small group of experts within an agency, there are many new tools that can democratize the decision by providing more information to the public. This chapter goes into more detail about the challenges of and new approaches to the utility/cost tradeoff. There are many lessons to be learned from past experiences.

MIT Press

Excerpted from How to Grow a Robot: Developing Human-Friendly, Social AI by Mark H. Lee © 2020 Massachusetts Institute of Technology.

Although I argue for self-awareness, I do not believe that we need to worry about consciousness. There seems to be an obsession with robot consciousness in the media, but why start at the most difficult, most extreme end of the problem? We can learn a lot from building really interesting robots with sentience, by which I mean being self-aware, as with many animals. The sentience of animals varies over a wide range, and it seems very unlikely that consciousness is binary—either you have it or you don’t.

It’s much more probable that there is a spectrum of awareness, from the simplest animals up to the great apes and humans. This is in line with evolutionary theory; apparently sudden advances can be traced to gradual change serendipitously exploited in a new context. As I’ve indicated, there are many animal forms of perception and self-awareness, and these offer fascinating potential. Let’s first try to build some interesting robots without consciousness and see how far we get.

Support for this view comes from biophilosopher Peter GodfreySmith, who studies biology with a particular interest in the evolutionary development of the mind in animals. He traces the emergence of intelligence from the earliest sea creatures and argues for gradual increases of self-awareness. He says, “Sentience comes before consciousness” (Godfrey-Smith, 2017, 79) and claims that knowing what it feels like to be an animal does not require consciousness. It seems entirely logical that we can replace the word animal with robot in the last sentence. Godfrey-Smith also argues that “language is not the medium of complex thought” (2017, 140–148, italics in original), which supports the view that symbolic processing is not a sufficient framework for intelligence.

In any case, it is important to recognize that the big issues in human life—birth, sex, and death—have no meaning for robots. They may know about these concepts as facts about humans, but they are meaningless for nonbiological machines. This seems to be overlooked in many of the predictions for future robots; systems that are not alive cannot appreciate the experience of life, and simulations will always be crude approximations. This is not necessarily a disadvantage: A robot should destroy itself without hesitation if it will save a human life because to it, death is a meaningless concept. Indeed, its memory chips can be salvaged from the wreckage and installed inside a new body, and off it will go again.

Consequently, such robots do not need to reason philosophically about their own existence, purpose, or ambitions (another part of consciousness). Such profound human concerns are as meaningless to a robot as they are to a fish or a cat. Being human entails experiencing and understanding the big life events of living systems (and some small ones as well), and human experience cannot be generated through nonhuman agents. If this contention is accepted, it should counter much of the concern about future threats from robots and superintelligence.

Two Nobel laureates, Gerald Edelman and Francis Crick, both changed direction following their prize-winning careers. Edelman won the prize for his work on antibodies and the immune system, and Crick was the co-discoverer (with James Watson) of the structure of the DNA molecule. Both started research into consciousness as a second career. Edelman experimented with robots driven by novel competing artificial neural systems (Edelman, 1992), and Crick looked for the seat of consciousness in the brain (Crick, 1994). They didn’t agree on their respective approaches, but their work, well into retirement, produced interesting popular books and showed how fascinating the whole topic of consciousness is. Despite their mutual criticism, their general goal was the same: They both thought that the circular feedback paths in the brain somehow supported consciousness, and they were looking for structural mechanisms in the brain.

I have already argued that sentient agents, like robots, need not be conscious, but they must be self-aware. In any case, it is a reasonable scientific position to start with experiments with models of self, self-awareness, and awareness of others and see how far the results take autonomous agents. Then the requirement for, or the role of, consciousness, can be assessed by the lack of it. This is not a structural approach, based directly on brain science as with Edelman and Crick, but rather a functional approach: What do models of self offer? How do they work? What is gained by self-awareness? What is missing from the behavior of sentient robots that consciousness could address?

MIT Press

Excerpted from Keep Calm and Log On: Your Handbook for Surviving the Digital Revolution by Gillian “Gus” Andrews. Reprinted with Permission from The MIT PRESS. Copyright 2020.

Who’s in the Audience? Would Your Followers Get Mad?

Broadcasting makes a lot of changes to how we communicate, especially when we’re suddenly speaking to more people who are different from each other. This often means changing the message we put out there, to avoid upsetting some people. And that can keep us from saying things we really need to say.

When people craft messages for a large audience—as they do in the advertising and PR industries—the tone they use in those messages is different than it would be for a small group. A person communicating to a large group of people will try to make their message easier for everyone to understand. They may work to avoid offending sub-groups in the larger audience. If they’re broadcasting in mass media, they’ll also work to avoid offending advertisers.

In television and radio, the need to be sensitive to the presence of children, elders, and potentially offended groups led to what are often called standards of decency. It’s one reason many countries set rules that tell broadcasters when they can and cannot air sexual or violent content, or use obscenities. In the United States, the Federal Communications Commission grants broadcast licenses. When the FCC fines broadcasters for nudity or cussing, they usually do so because of complaints from the public. So to some extent, offended members of the community drive FCC censorship.

The “wardrobe malfunction” incident at the 2004 Super Bowl (during which the singer Janet Jackson’s breast was briefly exposed by Justin Timberlake) triggered 540,000 complaints to the FCC. The Parents’ Television Council, a nonprofit watchdog agency, led the charge, demanding that TV be made safe for children. The FCC seized that moment to crack down on numerous other broadcasters, including fining the Fox network for running an episode of the animated series Family Guy that showed baby Stewie’s naked butt.

But knowing that angry letters from the public may be coming, broadcasters often self-censor even before the FCC comes down on them. Viacom and Clear Channel blacklisted Janet Jackson’s songs without being threatened. That’s when the radio personality Howard Stern, who had been fined plenty of times for indecency over the years, decided to flee from the terrestrial airwaves, later signing a deal to move his show to satellite radio. And the self-censorship effects went even further: after the wardrobe malfunction incident, ABC insisted it would air the movie Saving Private Ryan but cut its graphic war scene and expletives, even though the movie had previously been broadcast in full with no complaint from either the FCC or the PTC.

Who Are You “Singing” To?

Self-censorship based on “decency” limits the range of things that can be said in public. And there are parallels when we broadcast over social media.

My friend Steph explains that the difference between broadcasting and talking one-on-one is like the difference between singing on stage and singing a lullaby. You change how you sing, particularly how loud you sing. You wouldn’t do one in place of the other—you’d frustrate the audience by being hard to hear, or you’d wake the baby. 

This type of “using the wrong style at the wrong time” is obvious when people you know mix up their communication styles on social media. Most of us have probably cringed when someone revealed something too private in public, or treated a friendly conversation as a place to advertise. That cringe comes from our sense that different kinds of conversations have their place. Those who post something we find cringeworthy have not matched their way of speaking to our sense of who they’re speaking to. We may have called someone out for this, or heaven forbid, been called out ourselves.

Human beings have always had different ways of speaking to their peers, leaders, elders, or children. The tone of voice and the things that we say to these people just aren’t the same! And when someone uses the tone of voice with an elder that they would with a child, that elder might take offense. There might be consequences for the speaker.

If you’ve done the delicate dance of posting to Facebook—when you know your grandma, your drinking buddies, the kids you babysit, and possibly your next employer can all read what you post—you have a gut feeling for how to alter a message for broadcast. What you’re doing to walk that fine line is like what broadcast TV did for years: producing content for the widest possible audience.

Social media are really unusual in human history when it comes to social groups—they’re more like your city hall or national political conversation than they are like your club, church, nuclear family, or group of friends. They’re huge forums where everyone comes together, not small places where people with common interests have conversations.

Human societies have long had separate public and private spaces. (Here I use “public” to include the 150 to 290 person groups I mentioned earlier. When I say small or private groups, I mean small—less than a dozen people.) Sometimes “privacy” has been the difference between families and everyone else. Sometimes private spaces have been gendered, with women and men hanging out separately, or queer folks finding their own protective spaces to speak honestly with each other. Or they’ve been conversations only with friends of our own age.

Small-group conversations don’t have the same purpose in our lives as broadcasting information, or even reading the news or watching TV. They’re for soothing hurts, recognizing personal successes, assuaging doubts.

The huge forums of social media are often uncomfortable because we haven’t all agreed on what their purpose is. Scholars have called the lumping-together of everyone we know “context collapse,” meaning that the usual separate contexts we do things in (attend religious services, flirt with strangers, care for kids) are suddenly all mixed together. Awkwardly.

It’s important for us all to ask: Are technologies allowing us to speak to different groups of people we know separately, and speak to them in different ways?

One problem with broadcasting on social media is that the careful choices we usually make when speaking publicly can trample all over our need for private spaces with specific purposes. I’ll give you some examples of how this could play out.

We may go one of a couple ways when we broadcast and are at risk of offending people. The first way is the cringeworthy route. We all know people who just go the no-filters route and say whatever’s on their mind, no matter where they are or who’s around. In some cases, we might admire such people for saying what’s on EVERYONE’S mind—but we don’t always want to be them, because sometimes their mouths get them into trouble.

Or we may take the self-censoring route—the one that TV and radio have traditionally had to take—in broadcasting to a mass audience. And that’s where this gets problematic. Because social media put us out in public, we may not speak up when we’re struggling, worried that our next job or a college recruiter might see and limit our options in life. We may suppress new aspects of our lives we are exploring, like our sexuality or spirituality, which could benefit from talking to supportive others, because we are worried about the response from our church or family. Self-censorship hurts our capacity to grow and heal.

MIT Press

Excerpted from The Alchemy of Us – How Humans and Matter Transformed One Another by Ainissa Ramirez. Reprinted with permission from The MIT PRESS. Copyright 2020.

Long before the Great War, in 1895, science and magic were hard to separate. That year, Wilhelm Roentgen took a ghostly picture of his wife’s hand using mysterious rays that showed her bones. These invisible rays, later called X-rays, shot out of a contraption made of metal and glass that looked like something out of Dr. Frankenstein’s laboratory. Newspapers packed their pages with depictions of a person’s insides on the outside, and readers snatched up copies. Scientists were also enchanted by X-rays. Some of them wanted to know what else they could do. Others wondered where they came from. All these scientists understood that a battery attached to a stretched glass globe spawned a glowing stream called a cathode ray, and when this cathode ray collided with a piece of metal inside the globe, out came X-rays. Their thinking was that there must be more to these cathode rays. So while the whole world was wowed with X-rays, a few scientists were hoping to find the next great thing in cathode rays. Little did they know that this bright stream would explain how the world worked.

Cathode rays had been known for decades, but there was little consensus about their origin, and eventually the case went cold. With the renewed interest in them, scientists obsessed over every move of cathode rays, writing articles with reports of their behavior, though not yet knowing that cathode rays held a key to their science understanding. Locked in those cathode rays was the currency of all chemical reactions. Locked in those cathode rays was the answer to science questions from how toasters work to how planets were born. Locked in those cathode rays were the droplets that powered a river of modern technologies from televisions to computers to cellphones. Unbeknownst to these early scientists was that inside the cathode ray was a part of the atom that they didn’t know existed—the electron. But deciphering the puzzle of cathode rays required uncovering clues. Just as the popular character Sherlock Holmes used his intellect and his magnifying glass to solve mysteries, scientists too had to observe cathode rays under glass. For some scientists, this puzzle was too delicious to turn down, and Joseph John Thomson was one of them. It was this short man from the nineteenth century who would make the giant leap that made the technologies of the twentieth and twenty-first centuries possible.

Thomson’s potential in answering one of the biggest questions of his day seemed doubtful when he was fourteen years old in 1870. All he wanted to be was a botanist. As a small boy growing up near the city of Manchester, England, he spent all his pocket money on weekly gardening magazines. His father, a modest bookseller, wanted him to have a stable trade as an engineer. Being an engineer was good work, as Manchester’s textile mills turned American cotton into goods. To please his father, J. J., as Joseph John Thomson was nicknamed, attended Owen’s College in Manchester in 1870. But when his father died, J. J. scrambled to stay in school by winning scholarships. He entered Trinity College in Cambridge to study mathematics, choosing the beauty of numbers, instead of their utility, as in engineering. Walking on the hallowed grounds that Sir Isaac Newton strolled was an achievement for any son of a bookseller. But J. J. never fit in.

J. J. may not have felt at home at this old university, but his genius certainly was at home there. By 1895, Thomson was the thirty-nine-year-old head of the Cavendish Laboratory at Cambridge University, blossoming into an absentminded mathematics professor. His eyeglasses had two positions—one on his nose, which meant he was thinking, and the other on his forehead, which meant he was thinking more. He did not trouble his brain with worry about his appearance so his hair was long, his mustache overgrown, and his chin badly shaven. His brain was congested with abstract ideas, so his new research on cathode rays meant there’d be even less space to worry about ordinary things.

Uncovering the origin of the cathode rays was a perfect puzzle for J. J. because it challenged him by linking abstract ideas with observable events. Cathode rays shot from one electrical connection to another inside of a glass tube without air, and there were two dueling beliefs among scientists about how cathode rays moved in the world. One group thought that cathode rays were a wave that was a wrinkle in the ether. Others concluded that the beam was made up of small bits of particles acting together, like a migrating flock of birds. “Neither side was wholly right nor wholly wrong,” said J. J. There was evidence to support both ideas, but the cathode ray could not be both.

One definitive way to see if a cathode ray was a wave or a particle was to observe its dance with magnets. There was an old theory that said that if cathode rays fly undisturbed by a magnet, they are a wave; and if a magnet deflects the ray, they are made up of particles. J. J. wanted to test this theory and learned that fourteen years earlier, in 1883, another scientist performed this very same experiment. Cathode rays did not move when a magnet was nearby, supporting the wave argument. But J. J. thought there was something wrong with that earlier attempt. Scientific tools had advanced since then, and could draw more air out of a glass tube to better create a vacuum. A vacuum with less air was the habitat where cathode rays thrived best. So J. J., who believed that cathode rays were full of particles, wanted to repeat this old experiment using a glass tube with less air in it, made possible with an improved vacuum. J. J.’s mathematical genius, unfortunately, did not translate into manual dexterity. For such a small man, he was a Victorian bull in a china shop. When he visited his students in the laboratory, they’d wince when he offered help, and quickly tried to move fragile things out of his way. They took deep breaths when he sat on a lab stool to speak. Life was no better at home. J. J.’s wife did not permit him to use a hammer in the house.

J. J. needed help with his experiments and that help came from a former chemistry assistant, Ebeneezer Everett. While the name Ebeneezer conjures a miserly image, Everett was a dashing, mustached man, with cowboy good looks, who leaned a bit to seem less tall. Little is known of this Everett, except that he was a patient soul and a virtuoso for making laboratory glassware out of common soda lime glass into works of art that would have pleased a Murano glass master. Lab benches were full of Everett’s glass constructions, braced in place with wood brackets, with wires on every surface and sticking up into the air. Everett was the scientific brawn to J. J.’s brain. Starting in late 1896, J. J. wanted to make a cathode-ray Obstacle course to settle this wave/particle debate. Everett made a sophisticated glass bulb with pieces inside, reminiscent of a model ship in a bottle. On one end of the glass two metal pins stuck out that were attached to the ends of a battery to produce the cathode ray. Inside the glass, the cathode rays sprayed out in many directions like water out of a hose and were focused into a narrow stream, with two slits that acted like a nozzle. That beam then hit the interior surface of a round bulb, creating a green glow.

Cathode rays required that there be very little air inside the glass tube. “This was more easily said than done,” said J. J. To remove the air, Everett poured liquid mercury into a tower, which he connected to his glass bulb with a glass bridge. As the heavy liquid fell, it sucked air across the bridge from the glass bulb, creating a vacuum. Removing the air sometimes took most of the day, so Everett started in the morning before the hurricane in the form of J. J. Thomson arrived in the laboratory in the afternoon.

Only glass worked for these experiments. Copper would not do, nor any metal for that matter, for metals would bury the cathode ray. Wood or clay would not work either, for they could not hold a vacuum. Clear plastics hadn’t been invented yet. Glass was the best keeper of a vacuum; transparent, uninterested in conducting electricity, and malleable to an inventor’s imagination. But, mostly, glass was vital in science because it allowed scientists to do what they do best, which is to use their power of observation—and this was what J. J. excelled at.

Sometimes J. J. complained to his colleagues about his glassware. “I believed all the glass in the place is bewitched,” he said. Standard recipes did not yet exist for glass. Some parts of a glass tube were richer in key ingredients than others. To build with glass required compositions that were uniform all over, so that they would melt at the same temperature. And a glass piece divulged how well the bond was made only after many hours of work had passed. Sometimes glass whispered with a small air leak that there was something wrong, other times it screamed with explosions. Glass was temperamental, and it was up to Everett to tend to it like a newborn baby. In the summer of 1897, Everett completed J. J. Thomson’s obstacle course for testing cathode rays. He inserted two additional metal plates and attached them to another battery, creating an electric field, as a way to nudge the rays. As Everett turned the contraption on, J. J. saw that the cathode ray moved downward to the metal plate connected to the positive end of the battery. This told J. J. that the cathode ray was negative. Everett then put a huge horseshoe magnet around the center of the glass tube, and when he turned it on, J. J. saw that the cathode ray moved up, like migrating birds swept up by a strong wind. From J. J.’s mathematical calculations, written on the backs of random scraps of paper, he was able to deduce that the cathode ray was made of small bits that were electrically charged and negative. He calculated they were smaller than an atom, and were thus the tiniest part of matter yet discovered.

And when he and Everett repeated these experiments with different metal plates and with different gases inside the tube, J. J. saw that these same small negative charges existed in all materials. He called these bits corpuscles, but they would later be known as electrons. J. J.’s discovery changed the world, but he could not predict that it would. This small and odd man found the small and odd electron, opening up a door in science and expanding the understanding of matter. The discovery of the electron gave us clues about how galaxies and planets formed, because the exchange of electrons, in chemical bonds, explained how hot gases from the Big Bang coalesced into us. This discovery also revealed the basic building block of technology. With the electron, scientists would come to understand the workings of circuits, static electricity, batteries, piezoelectricity, magnets, generators, and transistors. With the knowledge of electrons, technology—and society—blossomed.

When J. J. Thomson was growing up, many inventions that we now take for granted did not exist. There was “no car, no airplane, no electric light, no telephone, no radio.” But the electrons in his glass, which made up electricity, would power all these machines as well as later developments such as computers, cellphones, and the internet. As smart as J. J. was, he could never have predicted that this abstract science would have practical implications. But it did, and it had many. With his discovery, humanity was thrust into a new age—an electronic one. None of these technologies, however, would have happened if it weren’t for the ability to see electrons in action. Our modern world was made possible by the ancient and old material of glass.

Excerpted from The Future of Brain Repair – A Realist’s Guide to Stem Cell Therapy by Jack Price. Reprinted with permission from The MIT PRESS. Copyright 2020.

The work of Gurdon, Thomson, and Yamanaka revealed something quite remarkable: if a cell can be induced to express the appropriate factors, then its fate can be fundamentally transformed. In the case of iPS cells, terminally differentiated cells—from blood, skin, or endothelium—were reprogrammed into pluripotent cells: that is, from cells with the most restricted of fates to cells with the most expansive. This was a shock to conventional embryologists, who had come to consider certain developmental steps irreversible. It was believed by many that once cells had been channeled during early development into one of the three primary germ layers (ectoderm, mesoderm, endoderm) then that step could not be reversed. Reprogramming destroyed that argument, but it raised an even more provocative question: if the correct genetic formula could be found was there any cell transplantation that could not be engineered?

The technique of iPS cell reprogramming takes a differentiated cell backward in development. From there, the cell can move forward again from the pluripotent state to become any of the various differentiated progeny to which such a cell would normally give rise (figure 12.1, below). The new question was: could reprogramming move a differentiated cell sideways; to another differentiated cell, for example, or a progenitor cell with a different fate? Could a fibroblast be turned directly into a neuron or a muscle cell? Or could it be turned into a neural progenitor cell or a bone marrow stem cell?

Remarkably, the answer to all of these questions turns out to be yes. As ever in science, there were straws in the wind long before biologists realized this was truly the case. Long before Yamanaka, a team in Seattle had shown that fibroblasts could be turned into muscle cells with a single gene.

The gene in question, MyoD, we now know to be a member of a group of transcription factors (bHLH genes) intimately involved in cell fate decisions in diverse tissues—heart, muscle, and brain. At the time, however, the molecular control of cell fate was largely unknown, and the existence of families of transcription factors was only starting to emerge as a consequence of the early genome sequencing efforts.

Colleagues, I recall, found this fate switch a troubling finding, but consoled themselves with the thought that these two cell types—fibroblasts and muscle cells—were actually pretty close embryologically, and anyway, strange things sometimes happened in tissue culture.

We have already met this phenomenon, “transdifferentiation”— the switching of cell fates—and noted that it has had a colorful history. While there were clear examples in vivo of cells apparently jumping from one fate to another, these were largely limited to “lower vertebrates” and involved closely related lineages. So, for example, if the limb of an amphibian is severed, cells within the stump dedifferentiate into progenitor cells (the “blastema”), which then regenerates multiple different cell types—muscle, dermis, bone—and thereby reconstitutes the lost tissue. In some species, heart cells (cardiomyocytes) can also dedifferentiate in response to damage, then re-differentiate following expansion to replace the heart tissue, and similar jumps have been observed in various tissues.

But these naturally occurring reprogramming episodes did not necessarily suggest that unrestricted reprogramming might be achievable experimentally. Following Yamanaka, however, a simple formulation emerged. If the combination of factors that prescribed a particular fate could be identified, then quite plausibly, expressing those factors robustly might make a cell adopt that fate. While the extreme form of this theory probably doesn’t hold up—that anything can be transformed into anything—nonetheless several quite remarkable steps have been demonstrated experimentally. Among them is the generation of neurons directly from fibroblasts.

The first demonstration of this came from Marius Wernig’s laboratory at Stanford. Their experiment reflected directly the approach that Yamanaka had pioneered. They sought the combination of transcription factors that would convert mouse skin fibroblasts directly into neurons, They found it required just three genes (Ascl1, Brn2, and Myt1l), and from this conversion emerged cells with all the significant properties of neurons: they grew a neuronal morphology, expressed the proteins that neurons express, formed synapses, and were electrically active. This was not, however, the first time that neurons had been directly reprogrammed from nonneuronal cells. Magdalena Götz and her collaborators had shown that transcription factors such as Pax6 and Olig2 modulated the capacity of glial cells to generate neurons.

But generating neurons directly from skin fibroblasts was an enormous leap in embryological terms: from a mesodermal end state (the fibroblast) directly into an ectodermal end state (the neuron), with no stem cell, or progenitor phase in between.

The neurons generated from this initial Wernig study, impressive though they were, were only characterized as generic neurons: no particular neuronal fate had been specified. The question therefore arose of whether specific populations of neurons could be generated. As we’ve seen, if the history of brain cell replacement has taught us anything, it is that we need the precisely correct neuron for each job. Several labs have now derived reprogramming formulas to generate specific neuronal populations, a number of which we’ve discussed in this book. For example, Ernest Arenas and colleagues at the Karolinska Institute in Stockholm have developed a protocol to generate dopaminergic neurons, while Andrew Woo and colleagues at Washington University in St. Louis have made striatal neurons directly from fibroblasts.

As well as indicating that clinically relevant neuronal populations are possible with this technology, these studies add a further wrinkle. It transpires that to achieve an optimal outcome, more than transcription factors need to go into the mix. At several points in this narrative, we’ve implied that cell fate can be determined by the correct combination of transcription factors. But as our understanding of cellular control mechanisms improves, we have discovered further cell components that participate in these processes. One such is noncoding RNAs.

For many years following the discovery of the genetic code in 1961, molecular biologists thought that the only essential role of DNA was to encode genes, which in turn encode proteins. Slightly alarming therefore was the discovery that only 1 percent or so of chromosomal DNA actually encoded conventional genes. The question became then: what is the other 99 percent doing? No less a person than Francis Crick is credited with concluding that it was probably “little more than junk.”

So the term “junk DNA” entered the molecular biologists’ vocabulary. But, of course, this had to be wrong. Were we seriously suggesting that a cell carried megabase upon megabase of DNA for which it had no use? Rather than deceiving ourselves by calling that 99 percent “junk,” we needed to discover what it was actually doing.

We now know that much of the genome (though still not all of it) encodes RNAs that do not encode proteins. These RNAs have a direct function, rather than just being vehicles for the transport of protein-coding information from the nucleus to the cytoplasm. That function, in many cases, is to regulate the cell’s translational machinery. They change the efficiency with which proteins are produced: proteins, which they themselves do not encode. Unsurprisingly therefore, they influence cell fate decisions, and can thereby influence reprogramming. In both of the direct reprogramming steps just cited, noncoding RNAs added to the mix improve the efficiency of the reprogramming steps.

This direct reprogramming has proven of interest to potential cell therapists for fairly obvious reasons. Instead of the laborious process of generating iPS cells, then taking them through a relatively long, complex process of differentiation, fibroblasts can be turned into the desired neuronal type in a single leap. There are, however, two issues with this approach, one practical and the other theoretical.

The practical problem is that, without the stem cell intermediate step, the possibility of expanding the cell population is lost. Neurons, as we know, are postmitotic: they don’t divide. With the iPS cell approach, each reprogrammed fibroblast gives rise to a line of iPS cells that can be infinitely expanded, ultimately giving rise to billions of neurons. But with direct reprogramming, each reprogrammed fibroblast gives rise to a single postmitotic neuron. This does not amount to many cells. A halfway house might be to reprogram from fibroblasts to neural progenitor cells, bypassing the iPS cell, but still giving rise to a dividing cell, which can itself then be expanded to give rise to many neurons.

Strategies are now in place to pursue this route. The theoretical issue relates to the mechanism underlying the direct reprogramming. Reprogramming iPS-style makes some sort of embryological sense. You make a pluripotent cell, then allow it to differentiate following the various embryological steps it would have taken in vivo. Direct reprogramming, however, makes no embryological sense. Nothing in nature, as far as we know, ever turns directly from a fibroblast into a medium spiny striatal projection neuron. This raises a number of questions regarding the veracity of directly reprogrammed change. Certainly, the reprogrammed cells have properties appropriate to the fate they’ve adopted, but have they abandoned all the indigenous programming that led them to their original fibroblast fate? This largely comes down to the epigenetic question we discussed earlier, and is the subject of current research.

#HashtagActivism: Networks of Race and Gender Justice
by Sarah J. Jackson, Moya Bailey, Brooke Foucault Welles

---

For people living in the margins of American society, Twitter’s ability to connect and amplify their voices offers unprecedented social and political opportunity. Hashtags like #BlackLivesMatter, #GirlsLikeUs, and #ICantBreathe have become rallying cries whose online impacts have resulted in real-world social movement.

But not every hashtag can be a winner.

#HashtagActivism explores how activists around the world have leveraged one innocuous feature in a social media app to not just call attention to their causes, but demand justice, and push for societal and government reform.

#CrimingWhileWhite: Consciousness Raising or Ally Theater?

In #CrimingWhileWhite tweets we see further complications of digital allyship emerge, but somewhat different from those of #AllMenCan. Some users engaging with the hashtag were clearly familiar with counterpublic discourses about discriminatory policing that emerged from #BlackLivesMatter networks and engaged in conversations and advocacy around these issues alongside the hashtag. These users produced a high co-occurrence in the network of #CrimingWhileWhite with #ICantBreathe and #BlackLivesMatter. While some such users were white, many were African American members of the larger #BlackLivesMatter interconnected counternetworks who found and shared #CrimingWhileWhite tweets via retweets or shares in which they added other racial justice hashtags to call their networks into the conversation. Rapper 9th Wonder, for example, became popular in the network because he actively retweeted white users’ #CrimingWhileWhite tweets to his hundreds of thousands of followers. Most users in the network, however, engaged with the hashtag but not at all with the themes or hashtags of larger conversations about racial justice online, reflecting either limited knowledge or limited interest in the larger #BlackLivesMatter and anti–police brutality networks.

While one reading of #CrimingWhileWhite tweets might locate them in the tradition of activist consciousness raising—that is, educating potential supporters and legitimizers of a movement about the political and collective meanings of their experiences (in this case, experiences with privilege)—other tweets suggest a navel-gazing that recenters (and borders on celebrating) whiteness. Within both the network and the discourse of the hashtag it is often difficult to tell whether particular users are, in fact, allies of the larger movement for Black lives and familiar with the Eric Garner case (which spurred the hashtag’s creation). Certainly, some users are, and the co-occurrence of such hashtags as #EricGarner, #ICantBreathe, and #BlackLivesMatter among some #CrimingWhileWhite tweets, along with some users’ connections to the larger #BlackLivesMatter twitter network, illustrates intentional allyship. Yet most #CrimingWhileWhite tweets do not contain these co-occurring hashtag nods to the larger movement, and most users do not connect with or engage larger questions and networks around issues of police brutality and racism. This observation illustrates (1) the limits of allyship hashtags in the larger digital public sphere, such that the hashtag may be adopted and put to use by individuals neither particularly invested in nor informed about a pressing political project, and (2) the limits of allyship tweets that do not amount to a direct call to action. While conversations about white privilege are encouraged by the hashtag, there was no call to action, and the question of what to do about that privilege, or, rather, with it, is not addressed.

Ultimately, #CrimingWhileWhite tweets perform two discursive functions, as memoir and to juxtapose experiences. Both these functions work to illustrate the benign or obliging role white Americans are accustomed to police playing in their lives—sometimes alongside implied or explicit commentary of the very different and more nefarious role of police for African Americans.

Memoir tweets in the #CrimingWhileWhite network recount compelling personal narratives about criminal behavior and interactions with police that are dramatic, emotional, and at times astounding. Every memoir is formulated similarly: the white person tweeting confesses to his or her own, often youthful, mischievous, or criminal behavior and then describes the fairly benign interactions with law enforcement that resulted (or did not result) and the few if any consequences the person ultimately faced. These tweets paint a picture of the normalcy of “acting out” and breaking rules for white people, particularly young white people, and highlight the fact that these young people are able to mature, grow, and live full lives thanks to an understanding society that does not criminalize them.

What these memoir tweets generally fail to do is directly engage networks focused on police brutality or suggest proactive ways white allies can intervene in the system whose privileges they describe. Rather, they seem to function as a type of catharsis, allowing users to acknowledge and confess to benefiting from white privilege without clearing engaging systemic battles with white supremacy.

Juxtaposition tweets employ discourse differently. These tweets generally include an image, meme, or link to a news story about a high-profile or notorious white person’s criminal behavior, and either directly or through implicit construction contrast laissez-faire law enforcement’s (and society’s) treatment of the person with the punitive treatment of African Americans. For example, we see the results of actions of actor Mark Wahlberg, former vice president Dick Cheney, and Ethan Couch, a young man who killed six people while drunk driving and who faced few consequences after his lawyer argued that Couch could not understand the consequences of his actions, contrasted with those of football player Plaxico Burress, Eric Garner, Michael Brown, Ezell Ford, and John Crawford.

In #CrimingWhileWhite juxtaposition tweets, white men who committed extreme acts of violence are frequently contrasted with Black men who were killed by police for minor infractions or simply by mistake. The implications are clear: white men face few, if any, consequences for bad, even extremely violent and criminal, behavior. Further, they are able to live not only productive but successful and powerful lives despite this behavior, while Black men are killed by police for far less.

In these tweets, white users move away from personal narrative, seemingly hesitant to directly compare their more generous experiences with the experiences of victims of police brutality and draconian criminal proceedings and preferring to use high-profile and egregious examples of other white people’s behavior. Here, users call out the unequal application of the law through proxy examples that remove their own experiences from the equation. These contrasting tweets do more to directly engage the larger network of racial justice tweeters by regularly using the #BlackLivesMatter and #EricGarner hashtags, yet they also seem to deflect personal accountability, making white people with extreme privilege the obvious “bad” guys and the police the obvious “racists,” rather than implicating themselves in the problem (or the solution).

While #CrimingWhileWhite tweets explicitly illustrate racial inequality in the legal system and white privilege through anecdotal examples, what is missing in both memoir and juxtaposition tweets is a call to action or acknowledgment of personal or group accountability and power to dismantle white supremacy. This makes the hashtag, while more organic and more popular than #AllMenCan, difficult to truly define as allyship, if allyship requires enactment of, or at the very least a call to action for the enactment of, behaviors that actively work to dismantle inequality. Of course, we cannot know what forms of activism users engage in offline, yet what we do know from #CrimingWhileWhite tweets is that personal accountability never became part of the discourse within the network.

#CrimingWhileWhite succeeded in expanding a narrative about double standards in the criminal justice system to a new audience, allowing white users and observers, some without connections to racial justice networks, to see the lack of severity in the policing and legal systems’ response to white people’s infractions as a common phenomenon, and begin to normalize a narrative about the differential treatment of African Americans. Yet at no point do tweets in the #CrimingWhileWhite network offer sweeping or systemic recommendations, solutions, or action responses to the phenomenon they identify, and thus they fall short of embracing the strategies for dismantling white supremacy that are popular in antiracist activist spaces. This reflects one of the greatest challenges in white allyship work in which white citizens see individual stories (such as memoirs and juxtapositions) and individual solutions (such as treating African Americans as they are treated or firing racist police) as more comfortable than acknowledging group complicity and responsibility for reimaging racist systems. Thus the consciousness raising and navel-gazing in the hashtag reveal the messy realities of online activism, in which ordinary users can take up hashtags and particular political narratives to be not either/or self-serving or reparative but both, in greater or lesser degrees.

Excerpted from #HashtagActivism; Networks of Race and Gender Justice by Sarah J. Jackson, Moya Bailey and Brooke Foucault Welles. Reprinted with permission from The MIT PRESS. Copyright 2020.

The Artist in the Machine: The World of AI-Powered Creativity
by Arthur I. Miller

---

Most of the time when we hear about AI, they’re taking our jobs or putting us in jail or inflicting some other autonomic horror upon humans. But there’s a second side to that AI coin. One in which machine learning algorithms show us skin suits the beauty of the natural form, even if it has been procedurally generated.

In this excerpt from The Artist in the Machine by Arthur I. Miller highlights the work of artist Eduardo Miranda. He’s melded the minds of a slime mold and a CPU to create, well, music.

Eduardo Miranda and His Improvising Slime Mold

What will be the central processing unit of the future? —Eduardo Miranda

Eduardo Miranda wants to shake up musical composition. At the moment, he is interested in central processing units (CPUs). In today’s computers, CPUs are silicon chips with circuitry that enables them to perform arithmetical, logical, and control operations. But supposing we go beyond silicon, beyond digital, beyond even a quantum computer? What about, for example, a bioprocessor that powers a biocomputer? Or a hybrid computer, powered by silicon plus a bioprocessor?

A bioprocessor processes biological material. Miranda’s chosen bioprocessor is a slime mold called Physarum polycephalum, the “many-headed slime,” a huge, yellow, single-cell organism packed with millions of nuclei. It is a mass of creeping, jelly-like protoplasm that is sensitive to light and spreads out over forest floors, eating fungal spores, bacteria, and microbes.

Physarum polycephalum has extraordinary electrical properties. Passing an electrical current through it makes it behave like an electronic component called a memristor. Memristors have a memory for current; their electrical resistance depends on the amount of current that has passed through them in the past. For Miranda, the most interesting thing about biological memristors is that they are not as precise as silicon-based digital ones. The mold’s electrical output is related to the input but in ways that can be hard to predict. He considers it a “creative processor.” Turn the pitch of a sound into electrical impulses and the slime mold will produce its own electrical response, which can then be turned back into music.

Miranda plays duets with the slime mold, which responds with enigmatic sounds produced on his piano. The piano is the interface between himself and his biocomputer. He entitled the first piece that emerged “Biocomputer Music: A Composition for Piano and Biocomputer.” There is no electrical connection between the slime mold and the piano. Notes played on the piano are converted into voltage oscillations that are fed through the slime mold in its petri dish. The vibrations cause it to change shape, producing energy that is turned into electrical energy and sent to electromagnets above the piano’s strings. The charged electromagnets attract the strings and make them hum, as if the mold is plucking them. Miranda has performed two pieces in collaboration with mold: “Biocomputer Music” in 2015 and “Biocomputer Rhythms” in 2016, in which the mold provided percussion.

Miranda grew up in Brazil and discovered computers as a teenager when his father brought home a Sinclair. He also studied the piano from the age of seven. He found it hard to juggle these two interests, AI and music. One hot and humid summer day, he visited the cool confines of the campus library and came across an article by Greek composer Iannis Xenakis, in which music was embedded in Venn diagrams, set theory, and logic, familiar to him from his studies in computer science. This, he realized, was the way to combine his interests in music and computers.

For his early research, he studied the way cellular automata behave in the Game of Life, a cellular automata system invented by the English mathematician John Horton Conway in 1970. In this game, a mathematical grid of cells follows simple rules about when a cell is “alive” or “dead.” The result is a riot of patterns. Besides its mesmerizing powers, the Game of Life turned out to have multiple unexpected uses—as a tool for exploring the evolution of spiral galaxies; calculating pi; investigating how ordered systems emerge from complex systems; and looking into why, in a multiverse scenario, only certain universes are capable of supporting life. Conway had hit on something universal.

Miranda was enthralled by the patterns and how they evolved. “This is very musical,” he thought. “It’s how certain composers explore variations on a theme.” He invented rules for how the patterns of a melody evolve over time, like the patterns in the Game of Life. “But,” he thought, “instead of visualizing these patterns as mathematics, how about listening to them?” This he accomplished by looking for patterns in musical ideas: phrases, rhythms, and harmonic structure. The machine encodes them and turns the results into lead sheets or MIDI files. Miranda had to curate the raw musical ideas to render them more playable. The computer, he feels, can only be “an assistant to generate ideas.”

Miranda notes that although neural networks can imitate musical styles, AI has not as yet produced any truly original compositions. This he feels is the next great step and perhaps can be accomplished with the help of his unpredictable slime mold collaborator. He is also interested in creating human-machine interfaces in which brain waves replace keyboards and voice commands to help the disabled express themselves musically.

In the introduction to his very personal book, Thinking Music: The Inner Workings of a Composer’s Mind, Miranda contrasts composing music using AI systems (neural networks) with algorithmic approaches, generating music from mathematical descriptions of fractals or of chaos. He writes, “Aesthetically, the algorithmic approach tends to generate highly novel and unusual music, whereas the AI approach tends to generate imitations of certain known types of music. … Both approaches have their own merits and pitfalls, but I tend to adopt the algorithmic one.” The biocomputer is one such innovative algorithmic approach.

Looking back over the interaction between music and technology in the twentieth century, he muses, “Cage and Stockhausen explored technology which they didn’t really understand. Now my generation understands technology and can use it to express our thoughts. It’s experimental music but we now understand what the experiments are telling us. So let’s make art now that is twenty-first- century music.” Miranda adds that he is not interested in reading about how a walk on a sunny day inspired someone to write magnificent music. Rather, he writes, “what I want to know is how he composed it. How did such inspiration become music?”

Excerpted from The Artist in the Machine: The World of AI-Powered Creativity by Arthur I. Miller (The MIT Press, 2019).

Sharenthood: Why We Should Think before We Talk about Our Kids Online
by Leah A Plunkett

---

Parents have been proudly posting pictures of their children’s development since the days of the derragotype, probably even before then if they could draw fast enough. But in the modern digital era — where Grandma and Grandpa are a Facetime call away, first words are live streamed on Twitch, and terabytes of baby pictures live in the cloud — it’s becoming increasing easy to overshare your kid’s daily trials and tribulations. In Sharenthood, author Leah Plunkett examines the impact of our ubiquitously connected world on children and the responsibility of adults to properly safeguard their children’s data.

In the excerpt below, Plunkett looks at an insidious form of “commercial sharenting” wherein parents exploit their children’s trust in order to gain internet notoriety and potentially impress a late night talk show host.

The second narrative type of commercial sharenting script is activities. The topics in this category run from arts and crafts to zany pranks, typically with a focus on doing activities with kids or guiding kids as they engage in such activities themselves. Some activities land on the quotidian. For example, the founders of WhatsUpMoms YouTube channel identify the search for travel tips for families as a key impetus to start creating content. Still other familiar activities include holidays, sports, and household decor. Sometimes these pursuits move beyond quotidian concerns into inspiration. Do more than get through the holidays: accessorize your home into transcendence!

Other activity categories are less common across households. Many moments in family life unintentionally create humor, as when you put laundry detergent in a dishwasher. Oops. But the laughs that bubble over from mistakes or spontaneous play are distinct from those that arise from elaborate plots to “put one over” on a family member. When that family member is a child, the prank may be no laughing matter. The dark side of the family prank space requires zooming in beyond the screenshot level. This side reveals how commercial sharenting can result in the total exposure of children at their most vulnerable. At its most extreme, such sharenting reveals to the world parental conduct that meets the legal definition of child abuse or neglect.

Recently, a court determined that a Washington, DC, area couple had neglected two of their children after a series of videos posted on the father’s YouTube channel, DaddyOFive, showed what to “most onlookers… looked a lot like abuse.” In an especially disturbing sequence, the parents spill disappearing ink in their son’s bedroom, swear and scream at him about how much trouble he’s in for the mess, then mock his justified indignation when he is told, “It’s just a prank, bruh!” This basic script repeats itself in a number of episodes: they put a child in an inappropriate or unsafe situation, capture his understandable emotional reaction, reveal that it’s “just a prank,” then document and ridicule his inevitable meltdown.

The court ordered two children removed from this family’s home and placed in foster care. The parents themselves had already suspended their YouTube channel, which had roughly three-quarters of a million followers. Viewers alerted authorities about the dangerous household. This development could suggest that the family’s YouTube postings, although a privacy intrusion for their children, were justified because they allowed outside eyes to witness the inner workings of this house of horrors. It could also suggest that the incentives to generate new and sensational content to capture viewers’ eyeballs contributed to this vicious and dangerous conduct in the first place.

Regardless of where you come down on these complex questions of causality and consequences, two general points about privacy and pranking are straightforward. First, after a prank is loose in the digital world, it is pretty much impossible to scrub it from the internet. The DaddyOFive YouTube channel is gone. Its content is the digital equivalent of real ink, however, rather than the disappearing kind. Its stain remains. The internet hosts perpetual reruns, whether the “actors” like it or not. The DaddyOFive content is readily available through other online sources, such as the YouTube channels of the viewers who have commented on it.

Even when that commentary is a respectful and thoughtful analysis of the “many ways to abuse your kids” and the reasons they’re all unacceptable, as one leading YouTube commentator put it, that commentator is still facilitating viewers’ access to the videos. Cody, the boy who was the butt of most of his parents’ so-called jokes, appears to have lived through a nightmare in the DaddyOFive household. In some ways, he will continue to live through one as long as that footage has an undead perpetual existence on the internet.

For Cody, decision makers about his current and future opportunities will not need a data broker to dig for or an algorithm to analyze intimate information about his childhood. His humiliation, fear, anger, and so much more are there in plain view. You would have to be heartless to hold any of his experiences against him.

But how about reasoning that goes something like this: “Of course, it wasn’t Cody’s fault, but given what we know about the potential for childhood trauma to have lifelong adverse impacts on survivors, maybe I don’t want to let my child have him over for a play date. Maybe I don’t want him in my class. Maybe I don’t want to give him a summer job.” Such questions are rational. They are also unfair to Cody. Depending on the role of the decision maker, they could shade over quickly into unlawful discrimination against him based on an assumption of disability. Perhaps more important, from a child’s perspective, they likely will make it hard for him to make friends and be himself, whoever that self turns out to be.

The second general point about privacy and pranking is that many kids today are subject to parental pranks. But there is a difference between so-called pranks that actually constitute abuse or neglect, like Cody experienced, and pranks that do not. A prank that is in poor taste or just not funny typically will be lawful. Today’s digital sharenting culture, however, does have an uncomfortable subplot of parental pranking to it even among commercial and noncommercial sharenters who avoid crossing the line into abusive or neglectful behavior.

Kids are natural comedic geniuses. Toddlers find it hilarious to repeat the old “throw the spoon on the floor, shriek for dad to pick it up, repeat” routine. Parents are also funny: they can make the spoon start to talk, flirt with the fork, and elope with the dish. Mazel tov! Maybe the family is the only one laughing, but it’s a spoonful of sugar to help real life go down. The sweetness starts to sour, though, when we get laughs at our kids’ expense rather than laughing with them or at ourselves. Take the annual trick or treat prank that late-night television host Jimmy Kimmel sets up every year.

Parents pretend they have finished all of their children’s Halloween candy, film their children’s response, and share the recordings digitally. The YouTube video of the 2017 “I told my kids I ate all their Halloween candy” challenge put out by the Jimmy Kimmel show has more than 2.8 million views. Kimmel gets contributions from sharenters everywhere. Spoiler alert: taking candy from a baby may be easy for the adults, but there’s nothing easy about it for the babies. These kids take it hard. Some of them have epic flipouts, and others struggle to remain calm while falling apart inside. The trick cuts deep, upending the immediate promise of Halloween mirth and the fundamental one of parental reliability. It generates a cheap and even sadistic laugh. That so many parents play along raises a disturbing question about the adult appetite for humor: how much of it is based on behavior that should be understood as bullying? It’s a loaded word, but cyberbullying might be the right term to describe the dynamics underlying certain instances of commercial and noncommercial sharenting.

In the last decade or so, there has been a growing focus by educators, lawmakers, and other decision makers on how to address bullying behaviors between youth, as well as to protect kids and teens from the harms that result. In many ways, the digital world has exacerbated these challenges and risks as children and adolescents engage each other around the clock across a range of devices and platforms. A common response by decision makers has been to pass new or update existing state statutes and regulations to require educator and law enforcement intervention when bullying occurs.

Let’s look at one such anti-bullying state law, which defines bullying as “a single significant incident or a pattern of incidents involving a written, verbal, or electronic communication, or a physical act or gesture, or any combination thereof, directed at another pupil which . . . causes emotional distress to a pupil.” The law specifies that bullying covers “actions motivated by an imbalance of power based on a pupil’s actual or perceived personal characteristics.” This law is binding only in the school context, hence the use of the term pupil. It is a law about how kids treat other kids.

Thought experiment: what happens if you swap in the word minor for pupil? The law then would prohibit a single significant incident that causes emotional distress to a person under age eighteen, including when that incident was motivated by an imbalance of power based on that person’s age. Publishing your children’s suffering—by taking Halloween candy from them, recording their reactions, and sharing the results with the world—seems to fit that adjusted definition. It is a significant incident that causes emotional distress to your child, however that distress is measured. An imbalance of power is inherent in the set-up of the incident. The parental role affords the adult “prankster” access to the candy. The child role puts the child in a place of dependence on the parent. What recourse does she have to get her candy back if her parent says it’s gone? The child role also virtually guarantees that the incident will garner a response that the parent sees as worthy of filming because, from a developmental perspective, the child is likely to have a strong and complex reaction to the “prank.”

Is it time to call in the parenting police? No, an anti-bullying law that covers parents and other adults won’t be written. Such a law likely would be unconstitutionally vague and overbroad. Especially as applied against parents, it could prohibit positive parenting conduct that keeps your child safe, like making your thirteen-year-old cry when you tell him he can’t drive your car because he’s underage. If the government proscribed even one “significant incident” of parental conduct that causes “emotional distress” to a child based on the respective parent and child roles, then the government would be intruding too far into constitutional protection for the liberty to parent and raise a family.

The rights to other adult-child relationships, like teacher-student or coach-athlete, are not entitled to the same level of constitutional protection as parent-child. However, these other roles do carry with them certain legal responsibilities that require adults to make decisions, based on the child’s age, that are necessary to keep them safe but may still cause the child emotional distress. Thus, an anti-bullying law that covers non-parent adult caregivers also likely would be too vague and overly broad to survive a legal challenge.

Although law enforcement won’t be opening a file for the case of the missing Halloween candy, we adults can and should still be thinking about the norms we adhere to in our daily lives. We don’t need a law to tell us that bullying our kids is wrong. We do need to think about how we explain the following to our kids: it is fine for us to take their candy, make them cry, film their crying, and share the video, but if they do the same thing to a younger pupil, they will get in trouble at school and perhaps with local law enforcement.

Is the right explanation similar to the one we give about drinking beer and driving cars? That explanation goes a little something like this: “You can’t do it now, but you can do it when you’re older.” Can we come up with a sound explanation here, one grounded in common decency and upholding the spirit of the anti-bullying laws our elected officials have passed for the schools that teach our children? If we can’t, then we should rethink the Halloween prank, both participating in it and watching it. More fundamentally, we should rethink our current acceptance of sharented “prankster” content by amateurs or professionals that makes kids the butt of jokes. There’s a lot more that is ghoulish than grown-up about it.

Excerpted from Sharenthood: Why We Should Think before We Talk about Our Kids Online by Leah A Plunkett (MIT Press, 2019)