February was the month I learned that asking “how could things be different?” can be a refrain of despair. It doesn’t have to be.

Six years ago, when I and my former PhD advisor learned of the MIT Media Lab’s secret relationship with Jeffrey Epstein, we chose to leave MIT and set aside the status that it conferred. Hardly a month since then has passed without someone asking me the “how could things be different?” They started as accusations of disloyalty, then pleas of desperation, and later something more. This month, amidst overwhelming revelations from Epstein’s partially-published emails, the question has been full of despair.

One reason the Epstein story feels so hopelessly oppressive is that the files were published by a powerful friend of Epstein who attended the same parties and allegedly had first-hand knowledge of his crimes. Epstein sought to control science with promises of money and influence that he sometimes delivered. His close friend and competitor Donald Trump seeks to control science by undermining public trust, censoring scientists, starving universities of public funds, and forcing university administrators to beg for help from the ultra-rich.

When people ask me my view of changes at MIT, reforms at Harvard, or consequences for Epstein’s enablers, I see why they’re stuck. As a behavioral scientist, I know that bureaucracy alone cannot ward off an abusive hierarchy of misogyny, racism, and power. I can’t sell them an answer simple enough for the cover of the pop science books so beloved at the Edge dinners. But I can offer three metaphors, grounded in the very science of networks that Epstein was so obsessed with owning.

Men like Epstein seek to control women through what social scientists call direct reciprocity. In cases of power imbalance, what seems like generosity is a strategy to narrow a person’s life by providing, controlling, and expecting compliance in return. The jobs Epstein gave his victims were designed to trap them, not support them. Scientists draw this structure as a loop between two people, and it rightly looks like a manacle.

Ending this kind of relationship is costly and can be dangerous under threats of violence. In conditions of extreme inequality, too many people owe too much to a wealthy few. And everyone who owes something to abusers or their enablers becomes a potential accomplice. When people appealed to loyalty in their fearful arguments with me, I saw how far the chain can reach.

Epstein offered powerful men the gift of indirect reciprocity. In this web-like structure, givers contribute to a public good in exchange for reputation. Epstein’s emails show how he coaxed prominent scientists and business leaders with promises of funding and introductions. He then traded on the reputation they conferred to expand his crimes and web of influence. This explains why so many luminaries have blamed others for convincing them to join Epstein’s network, and also why the sticky strands, once revealed, have been so hard to wash off.

The leaders I trusted who were secretly in Epstein’s circle loved to talk about networks as bridges to opportunity and resources. They had a point. But bridges are powerful tools for gatekeeping. And the more embedded someone’s network becomes, the harder it is to join or to leave.

I will never forget the desperation, courage, and solidarity of students who had come to college for new connections and somehow found themselves navigating a tangle that had been dropped on them. When Ethan and I left, we had no idea how large a tear we were making in the fabric of our personal and professional relationships. For years after, I would briefly think of someone who could support one of my own students before remembering, like an amputated limb, that I didn’t know where the relationship stood. I expect some of them have felt the same way.

The manacle and the net remind me of the false sense of determinism promoted by the pop science books that Epstein supported. Luminaries sell us narrow stories about the power to control ourselves and each other when the true beauty in the study of humanity is in variation.

As the communities behind women’s shelters know, chains can be cut. Knots can be untied. Nets can be remade into something new. Because people also organize a third kind of reciprocity. In upstream reciprocity, people contribute without the expectation of benefits in return.

We see upstream reciprocity when someone does the thankless work to stand up for gender equality or support a survivor to seek justice. I saw it in the work of an advisor and his counterparts who ensured that his students graduated after he left. I also see upstream reciprocity in the work of leaders — let’s admit they’re mostly women — who are given the job of picking up the shards of shattered institutions.

But upstream reciprocity doesn’t just take from people — it can also give. When courts try powerful people for their harms and send the damages to public institutions, they sever the bonds of obligation and reciprocity that often follow money. When the good neighbors of a democracy tax the rich to support schools, community colleges, and universities, they untie money from abusive oligarchs and bind the scientific endeavor to the common good.

When social scientists draw upstream reciprocity, we use the shape of a tree.

In a time of so much inequality and mayhem caused by powerful men who dominate public attention, despair offers the false choice of which powerful man to offer favors for protection. But for every person forging manacles or tying nets, there are a hundred people planting trees.

Now when people ask me how things could be different, I have an answer. It’s not my role to say what you or your institution should do. But I have found a home among some great gardeners. And we are planting trees.

A note to readers: a metaphor is also a kind of seed, whose power begins when you let it go.

I will not be responding to comments, critiques, or advice about this post. I wrote it for those who told me they were struggling with the deafening silence in academia in February 2026, and it is all that I have to say for now.

Scientists are sometimes blamed for delaying accountability and public knowledge on clear harms and injustices. In some cases the blame is fair — when a powerful actor demands impossible scientific evidence on a harm that everyone can see.

But if science is a source of delay, why rely on it at all?

I’m a computer scientist and behavioral scientist at Cornell University who works together with the public to study the impact of technology on society, and work together to test effective interventions for a safer, fairer digital environment. I’ve worked alongside the public to investigate issues including online violence, algorithmic discrimination, and youth mental health.

In every project, community partners and I have a conversation about the norm in science — where we commit to following the evidence wherever it leads. I am always grateful and inspired when people work with us when they are being ignored by powerful institutions and are hoping that science will validate their claims. A successful partnership also requires us all to prepare to learn something we didn’t expect.

I also have these conversations with students at Cornell, in the class I teach on investigating AI failures. This week, I introduced students to one of the most horrific cases of unscientific evidence run amok: research on the causes of sexual assault.

In the US according to the CDC, nearly half of women and more than 1 in 6 men have experienced some form of contact sexual violence in their lifetimes, with more than 1 in 5 women experiencing completed or attempted rape. Given this endemic, society-wide harm, many schools and colleges put significant resources into understanding sexual assault and trying to prevent it.

Unfortunately, many interventions on sexual assault are based on a single study with a small, non-probabilistic sample that promoted the “serial rapist” model. The flawed paper claims that a small minority of atypical people are responsible for over 90% of sexual assault on campuses- that it is the result of “a few bad apples.” If you have ever taken a bystander training, you have been trained in the serial rapist model and have probably heard this number.

The problem is that the “bad apples” theory is not accurate. Multiple large-sample longitudinal studies have not found evidence for the serial rapist hypothesis. While some people are repeat offenders, they find evidence that sexual assault is more endemic to society and perpetrated by a wider range of people. Yet findings from that single, flawed 20+ year old study refuse to go away. As a result, many people and organizations have held onto a fundamentally mistaken framework, spending countless millions of dollars and leaving people unprotected for decades.

How can a debacle like this happen? Psychologists Ana Gantman and Elizabeth Levy Paluck recently published a paper finding that the serial rapist model, even though it’s not true, is associated with people’s belief in a just world and related beliefs. Confronting the truth would be too painful, forcing many people to rethink assumptions about our communities. So people latch on to bad science and throw millions of dollars at misplaced efforts that provide cover for the ongoing scourge of sexual assault.

Confronting the truth would be too painful, forcing many people to rethink assumptions about our communities. So people latch on to bad science.

One of the things I admire the most about Gantman and Paluck’s work is that they didn’t stop after pointing out the problem — they have been tireless at working with communities to test interventions to change the climates that enable sexual assault. By shedding light on misconceptions, they have created new space for solutions.

On its best days, science helps us ground our beliefs in what is observably happening in the world. Whenever someone pressures me to publish data that confirms my assumptions, especially when I feel a sense of urgency and moral outrage, I think about the researcher whose quick, unreliable study on sexual assault set back public health by decades.

Yes, it’s bad when slow science delays a search for solutions. But it’s sometimes even worse when fast, unreliable science creates a false trail for well-meaning people by misleading the public about harms that need urgent solutions.

It can feel like a miracle when scientists and hurting communities find common cause in a search for the truth, and I’m filled with wonder and hope every time it happens.

Further reading:

• COMM 4940: Investigating AI Failures
• Gantman, A. P., & Paluck, E. L. (2025). What is the psychological appeal of the serial rapist model? Worldviews predicting endorsement. Behavioural Public Policy, 9(2), 461–476.
• Pierson, E., Li, S. (2015) A better way to gauge how common sexual assault is on college campuses. The Washington Post
• Rosenberg, M., Townes, A., Taylor, S., Luetke, M., & Herbenick, D. (2019). Quantifying the magnitude and potential influence of missing data in campus sexual assault surveys: A systematic review of surveys, 2010–2016. Journal of American college health, 67(1), 42–50.
• Porat, R., Gantman, A., Green, S. A., Pezzuto, J. H., & Paluck, E. L. (2024). Preventing sexual violence: A behavioral problem without a behaviorally informed solution. Psychological Science in the Public Interest, 25(1), 4–29.
• The Monument Quilt — oral histories from the public healing space by and for survivors of rape and abuse established on the National Mall in Washington D.C. in the summer of 2019, and recorded by StoryCorps.
• Matias, J. N., & Price, M. (2025). How public involvement can improve the science of AI. Proceedings of the National Academy of Sciences, 122(48), e2421111122.
• Orben, A., & Matias, J. N. (2025). Fixing the science of digital technology harms. Science, 388(6743), 152–155.

“Every being cries out silently to be read differently.”
― Simone Weil, Gravity and Grace, quoted by Francis Su in Mathematics for Human Flourishing

In the last twenty years of my journey as a scientist and organizer, only three pieces about mathematics have brought me to tears.

The first was a study by Patrick Ball on State Violence in Guatemala, a statistical analysis that helped me understand the genocide and civil war that moved my indigenous father and American mother to stay in the US to escape the violence. The second was Francis Su’s beautiful essay “Mathematics for Human Flourishing.” Su tells the story of an incarcerated person who finds refuge and hope in mathematics — and asks us whether we are willing to consider such a person a mathematician.

The third work on science that brought me to tears is Voices in the Code by David Robinson, a book that traces the remarkable debate over how the US allocates donated kidneys to people who need them. Robinson tells us the story of Clive Grawe, who traveled hundreds of miles to join doctors, surgeons, and statisticians to debate the finer points of allocation algorithms. Unlike those professionals, Clive was an expert of another kind — his daughter desperately needed a kidney and her livelihood depended on what algorithm would be chosen.

I got to reflect on these stories this week when Megan Price and I published a new article in the Proceedings of the National Academy of Science on “How public involvement can improve the science of AI.”

All of these works share a common thread: the idea that science can, if we let it, help society negotiate the hardest questions we face when we respect the expertise of people’s life experience.

Megan and I open our article with the story of nurses who went on strike in search of a seat at the table of AI evaluation. Our article asks the question: can public participation in AI evaluation make the science better or is it a necessary compromise of quality for the sake of negotiation and democracy? Megan and I hope our summary of over 170 articles helps answer this question for scientists, advocates, funders, and policymakers (you can read the article here).

It’s not common to publish narrative in peer reviewed articles, but our editors and reviewers at PNAS let Megan and me tell stories about communities whose expertise was respected in the scientific process, and what we learn from those stories. In each case, we describe how people’s lived-experience expertise made the science better. I hope we did justice to their stories.

To me, this article is so much more than just another academic article. It distills a decade of trying to do good science together with communities. It also brings me full circle with one of the organizations that motivated me to become a scientist in the first place.

My co-author Megan, you see, is Executive Director of the Human Rights Data Analysis Group (HRDAG), the organization whose analysis brought a younger me to tears so many years ago. In 1999, U.S. president Bill Clinton issued a formal apology for the US role in the Guatemalan violence toward indigenous people, citing the report that HRDAG developed with input from people affected by the violence. It was an extraordinary moment in world politics, and it gave me a glimpse of how mathematics can sometimes be a shared language for important conversations about truth, justice, and shared interests.

I’m so grateful to Megan for a year and a half of collaboration, and to the peer reviewers who validated our article for one of the world’s leading general scientific journals..

Some readers will already be thinking about the shortcomings of science in society and all the ways that power and money and misunderstandings often distort the common good. I hear you.

All the same, I hope our article inspires hope by showing numerous cases where people faced tremendous complexity and conflict, finding shared understanding through a commitment to the truth, however begrudging.

And perhaps Megan and my article might someday catch the eye of a young person who’s asking how they can pursue truth with care and rigor in a world that so desperately needs to keep this vision alive into the next generation ❤

On August 1st, I’m riding my bicycle 10 thousand meters of elevation in under 36 hours to find out. As I ride the height of Mount Everest from my home in the Finger Lakes, I’ll be guided with a mixtape from an extraordinary group of South Asian bloggers, translators, and activists, in a fundraiser to support Global Voices (I’m on the US board).

There was a time when people thought journalism could query and report on a world of social knowledge at scale. That time is over, as Henry Cooke & Ian Forrester recently wrote in a report for the BBC with projections on the social Internet. Instead, we need to rely on the basics of good journalism: people with trusted relationships who can understand, explain, and translate specific contexts to a wider audience.

That’s the idea at the heart of the work of Global Voices, a network of thousands of writers, editors, and translators who work in over 30 languages to report on stories around the world, translate those stories, defend free speech online, and empower underrepresented communities to tell their own stories.

This coming weekend (August 1–3), I’m going to be riding an “Everest Roam” on my bike, climbing over 32,000 feet in elevation across Central NY in under 36 hours to raise funds for Global Voices and for the simple idea that a better world depends on human connection and human understanding. If that’s something you believe in too, I invite you to donate.

a better world depends on human connection and human understanding

The sport of endurance cycling loves a challenge, and one of its biggest challenges is Everest Roam, where riders climb more than 10,000 meters in elevation (32,809 feet) in less than 36 hours, riding a minimum distance of 400km (248 miles). In my wildest dreams, I would love to collect a dollar for every meter in this 10,000 meter ascent.

To illustrate the beauty and glory of the Global Voices community, the Nepal team is creating a playlist for me to learn more about Himalayan culture on the road. I’ll share it as soon as I have it!

Folks who subscribe to my Medium posts will know that my cycling Everest Roam attempt is the latest in a series of journeys of understanding I have undertaken while coming to terms with a respiratory condition that changed my ability to travel widely.

In 2022, I rode a “Pilgrimage for a Million Lives,” 190 miles representing 12 inches for every COVID death. In 2023, Ivan Sigal and I rode through California’s Central Valley in the footsteps of the 1966 Farmworker March to learn about the future of food and the environment. In 2024, I designed the Maple Bicycle Adventure to search for hope on crises that feel impossibly big to change. And this summer, I rode across the Great Divide from Colorado to Utah to re-think the true power in people’s relationship with technology.

In case you’re wondering, training is going well! Saturday, July 26th, I got to ride around 16,000 feet in 12 hours while listening to Global Voices Executive Director Malka Older’s delightful novella The Mimicking of Known Successes and having a long, rambly phone conversation with my partner about Petrarch’s literary ascent of Mount Ventoux.

Where do you turn when you experience something so extraordinary that you lack the vocabulary to explain it, even to yourself?

I felt that way last week when pedaling my bicycle with a week of luggage over America’s Continental Divide at eleven thousand feet. Surrounded by mountains, snow, and streams, I remembered the Puritan colonist William Bradford’s first words upon sighting the continent: it was “a hidious and desolate wildernes, full of wild beasts and wild men… [with] little solace or content.”

In 2025, Bradford could easily be mistaken for describing our digital environments, on topics from smartphones to AI. Like the Puritans who left religious persecution for the cold and starvation of a different place, we are also told by venture capitalists that the risks of being left behind are greater than the dangers of systems they cannot reliably manage. When people face pressure to populate the digital frontier or else, it’s reasonable to worry about the dangers, especially when technologists downplay the risks. And it’s not surprising that fear can drive us, like Bradford, to ideas that de-value our fellow humans as “wild men.”

many people would agree that our digital environments have become ugly and risky

How can we turn this around? As I stopped my bicycle on the Fremont Pass and enjoyed the view in the cold morning air, I thought of all the people who flock to the Rockies out of love rather than fear. Centuries after Bradford, Americans now find peace and inspiration in a landscape that the colonist viewed as hideous and desolate. As Phil Ochs wrote in the 1964 song “Power and the Glory” (famously covered by Pete Seeger):

Here is a land full of power and glory
Beauty that words cannot recall

Somehow over 400 years, the inheritors of colonization have learned to love the glory in nature and not only fear its power. We now spend 55.6 billion dollars annually to enjoy and preserve our National Parks alone. Nearly three million young people each year participate in Scouting America and the Girl Scouts, which encourage adventure alongside care for nature and community. Love for the land, in all its complicated history, is one of the things that unites Americans, even if we sometimes disagree on how best to do so.

Looking down from Fremont Pass at the fourteen thousand acre molybdenum mine that probably supplied the ore for my bicycle frame, I thought about how the Pilgrims and today’s Americans manage fear by seeking to control and exploit what we fear. In this view, something is to be enjoyed when it has been tamed or put to work. Yet even as we try to calm our fears by controlling nature, we have come to mourn how much gets lost along the way.

Fighting fear with control is happening everywhere right now as governments consider banning smartphones and social media for young people. In the American West, engineers tamed the landscape with dams, mines, and underground nuclear detonations. These projects tried to control nature for safety and profit, often before the consequences were fully understood. Scientists are now being asked to do the same for technologies of human thought and connection, often before the science is ready or fundamental freedoms have been taken into account. Comparing the history of natural resources to the tech industry, I see a common thread in the drive for unbridled development — which heedlessly pursues growth and responds to public concern with more of the same.

As a scientist-organizer, I understand the value of fear, anger, disgust, and outrage. In healthy doses, these powerful emotions can motivate change. But they also burn out quickly and can leave us desensitized — they’re an unsustainable fuel for care and maintenance over the long term. That’s one of the great lessons to learn from broad-based initiatives across the conservation movement, which balance outrage with love, wonder, and community.

Standing with my bicycle at the Great Divide, I faced a choice: to continue over a legendary, stunning climb to Independence Pass or take a shorter route to Utah through the canyon at Glenwood Springs. This shorter route would give me several hours to support a group of deeply-committed parents who are working for safer social media in the long term. Weighing this decision, I remembered the next lines in the song by Ochs:

Her power shall rest on the strength of her freedom
Her glory shall rest on us all.

In this deceptively simple song, Ochs is asking us to look on our fellow humans with the wonder and love that we have learned to offer to nature. If we can do that, the song implies, we can find a way to maintain and protect people with the same care that many of us offer to our rivers, forests, and mountain peaks.

As a scientist who works alongside the public to study and improve our digital environments, I regularly get to witness the power and glory of people working together out of love. To name a few that CAT Lab has worked with, I’m inspired by Jefferson Kelley, whose love for his son led him to transform how Reddit’s algorithms depict Black fathers. I’m also grateful to the hundreds of volunteers who helped sustain public knowledge by sending thank-you notes to thousands of Wikipedia editors — as well as clean up vandalism on the site. And I’m deeply moved by the enduring commitment of parents who have campaigned for better science on social media and AI after the deaths of their children.

No complex system, from waterways to digital safety, can be protected with a single tool, whether it’s the law, science, or design. And while fear, outrage, and a demand for justice can often get us out of our seats, nothing healthy can be cultivated over the long term when those emotions replace rather than motivate cooperation.

Instead, movements built on wonder and love have the capacity to become greater than the sum of their parts. According to the U.S. Census Bureau, nearly twenty-six thousand people were working in conservation science and forestry in 2022. That’s less than one tenth of the three hundred thousand volunteers who contribute more than 6.5 million hours of service to the park system annually. We desperately need the professionals who maintain our National Parks, and many Americans are angry that so many have been fired without cause this year. Also, volunteers are stepping up to support the land we love at a time when these public goods are especially at risk.

Balanced on my bicycle with the chance of a once-in-a-lifetime view just fifty miles away, I took the shorter route instead. Independence could wait for another day. I didn’t want to miss the chance to witness and support the glory of human love.

In 2025, many people would agree that our digital environments have become ugly and risky. For some the solution is to escape. For others, the answer is to control. There’s a third way that many have already taken: a path of cultivation, science, cooperation, and love. If you haven’t tasted this yet, I invite you to experience and create this glory together.

When I was a gradstudent, I was inspired by Public Lab’s work to collect environmental data on the BP Deepwater Horizon Oil Spill despite corporate and government resistance. So when founding the Citizens and Technology Lab to study our digital environments, I decided to prioritize industry-independent community science. By organizing alongside affected communities, Public Lab circumvented restrictions on science. We do the same.

For a snapshot of my thinking then, see this 2014 Atlantic article that launched the lab.

Years later when tech companies including Twitter and Meta cracked down on research in what Deen Freelon calls the “post-API age,” CAT Lab had zero disruptions to the active research infrastructures we use to study online behavior at scale. We and our community partners continued our research unperturbed. This freed us to co-organize with many other, more-disrupted researchers to create the Coalition for Independent Tech Research.

When critics question the reliability or sustainability of community/citizen science, they often have in mind an idealized policy environment where well-resourced institutions enable and support science. Where those institutions exist, they are indeed valuable. Also, those idealized situations are sadly often the exception, and scientific institutions can also become single points of failure.

As we head into a period of concern about the resilience of government-supported data, I’m glad to see so much effort to protect and sustain official data sources. I’m also grateful to all the community/citizen scientists out there who will keep collecting data to advance the public good whether or not governments or institutions have our backs. We’re going to need this work more than ever. ❤

What does a trustworthy tech governance system look like? Last month I saw the most democratic diagram I’ve ever seen — the schema for who gets notified when a U.S. Environmental Protection Agency air monitor is re-calibrated.

As some of you know, I do environmental pollution research on the side to manage my health and give back to society. To manage my everyday life, I get to analyze official EPA air sensor data, learning how the EPA does science — which is SO COOL.

The US National Ambient Air Quality Standards (NAAQS) is a remarkable public data system from the 90s that collects and publishes data from over 5,000 sensors across the country. Some of these are permanent sensors; others are mobile sensors placed at the fence-line of an industrial facility. These measurements support:

• evaluation of whether a geographic area or industrial facility is in compliance with the Clean Air Act
• emergency response & public knowledge during events like wildfires or dust storms
• evidence in legal cases

Real-time pollution monitoring is hard to do reliably. Last month, for example, I noticed that a little moth had grown a chrysalis inside one of my personal sensors, distorting the pollution measure. I had to decide — do I throw out a month of data or add a “moth adjustment” to my records? Complexities like this one are why the the EPA only has around 5,000 active monitors: precise measurements are expensive and monitors need regular testing/maintenance.

How should democracies handle real-time data? Timely air quality data is a matter of life/death, including mine. It shapes millions of dollars in industry costs. It also directs tax money (via government cleanups). Since we live in a democracy, changes to the Air Quality Index (for example) or re-calibration of a sensor can’t just be done by scientists when they see a problem. Changes need to be to be debated.

For example, last year the EPA adjusted the math to calculate the Air Quality Index (after an extensive public consultation). This was great news for me, because I’ve been puzzled by my own health effects at concentrations of pollution that the EPA used to consider safe. Then in 2024, the EPA adjusted the Air Quality Index to better match the state of public health science. Like everyone else (I hope), I had to spend a weekend changing the code of my sensor network to re-calculate all current, future, and past measurements.

Not everyone thinks that updating the air quality calculations was a good idea, even if it will save lives. Lobbyists for manufacturers said that it was too burdensome. Attorneys general of 24 states agreed. After they lost the debate in the public comment period, they sued the EPA to return to standards that would allow higher levels of pollution. That’s their right. In a working democracy with effective governance, legal attacks on a regulation are positive signs that the regulation is consequential.

For public debate in a democracy to work, people need to know about the debate. And the EPA has built that debate into their air quality monitoring software. This diagram is the schema for defining who needs to be involved and/or notified when EPA scientists propose a corrective action to a sensor measurement. It includes federal and regional authorities, tribal nations, and many other entities who need to know about (or approve) small changes with big potential consequences.

I understand — many people will look at this and see needless bureaucracy. Some industry lobbyists will claim publicly that it’s a needless maze while seeking privately to make it as complex as possible to delay accountability. Advocates will rightly decry how slow it is to wade through endless legal battles for decades while whole generations are facing irreversible damage to our health. And they’re right. At the same time, this messy, complicated democratic system, supported by reliable sensor measurements, prevents 230,000 deaths per year from particulate pollution alone.

Ultimately the Clean Air Act is democracy — built right into science and the software of regulation, and it’s beautiful.

If you want to learn more about the science, politics, and software of environmental data, I encourage you to follow the work of the Open Environmental Data Project. If you want to see new innovations in how advocates are using science to create change, check out the Central California Environmental Justice Network, who I have written about for Global Voices. If you’re curious about inequality and injustices in the sensor network, a good place to start is this excellent ProPublica report on the politics of where sensors get placed (and where they don’t). There’s a new study in JAMA about this as well (Thanks Neil!)

And if you want to meet other community/participatory scientists who are working to broaden access to data and justice, I hope to see you at the annual conference of the Association for Advancing Participatory Sciences in Portland this May!

On a hillside in Tully New York, Nate and Cristy Williams convert thousands of gallons of sap into syrup each year. They also produce maple’s most important missing ingredient.

Last March when I pedaled my bicycle through fifty miles of snow to search for that ingredient, I was also looking for something that often feels even harder to find. I was searching for hope on a crisis that feels too impossibly big for any one person to make a difference.

Anyone who’s ever walked, skied, or bicycled for hours in the snow knows the feeling of entering a crack in history — surrounded by landscapes that could have been painted a century ago. In the Finger Lakes where I live, white-coated fields and forest hills are punctuated by the polka dots of red barns and white farmhouses sending woodsmoke into the air. Rolling through the landscape on two wheels beneath several layers of wool, it’s easy to imagine a world without cars, the Internet, or fast food. If this feeling were a timezone, I would call it Maple Syrup Time.

The Tully Valley offers the kind of views that make Central New York an endless painting so beautiful you would call it kitsch. Just a few miles south of the Erie Canalway Trail near Syracuse, winding roads follow the Onondaga creek along forests and farmhouses, through the Onondaga Nation, and into Onondaga Lake.

Beneath the snow are fields. Beneath the fields are fertile soils. And at the southern end of the valley, under the clay and sand and shale is a void.

Next time you eat a piece of bread or bake a cake, take a look at the empty spaces that make it so light and moist. Those voids are produced with the help of baking soda, a compound that reacts inside the dough to create bubbles of gas much faster than yeast. In the 19th century, bicarbonate of soda was also being used to make glass, soap, and wool. So when Belgian chemist Ernest Solvay met an engineer from Syracuse with a tip on deposits of brine and lime, they decided to start a mine along the Erie Canal to meet the demand.

As my wheels crunched over the snow on Solvay Road, I thought about the century of brine mining that hollowed out the land beneath me in the southern Tully Valley. Starting in 1890, the Solvay Process Company injected water into the bedrock 1200 feet below. This water dissolved rock salt buried in the floor of Paleozoic oceans, removing over 96 million tons of sodium chloride over 90 years from the valley’s southern end.

Alongside industrial pollution the company deposited downstream, Solvay engineers recorded patterns of subsidence starting in the 1930s: sinkholes the size of football fields, hundreds of cracks in hillsides above the brine fields, and a general reduction in land-surface elevation of 40–70 feet in and near the sinkholes. Geoscientists cannot conclusively say how much the mining contributed to catastrophic cracks and land subsidence, but the mine closed in the early 1960s, and the western brine field closed in the 1980s due to the high costs and concerns about the environmental damage.

When the US Geological Survey investigated bedrock fractures a foot or more wide and 30 feet deep near the Tully brine fields in 2006, they noticed something remarkable: a network of maple roots connected both sides of the broken earth.

One maple germinated in the center of the crack in 1927 just as it broke open. For the next seventy-nine years as the rift widened, the soil around the tree’s roots fell away, leaving the tree to grow across the chasm below. No tree can hold together a fractured land, but roots can still support life by drawing water and nutrients across the divide.

If you follow Onondaga Creek upstream from the brine fields and past a series of waterfalls, you will reach Woodmancy, a winding forest road that hugs the hills a mile west and seven hundred feet above the old mine. These hills are covered in a forest of maples. Below ground, the trees share bonds of roots that hold the topsoil firm. Above ground, they are knit into a network that converges only in winter and disappears each spring: bright blue tubes that lead to the farm I cycled six hours in the snow to see: Dutch Hill Maple.

Walking into Nate and Cristy’s sugar house is not unlike entering a historic chapel, if chapels were covered in steel siding painted green.

The atrium is a shop, its shelves full of relics in the church of sugar: syrup bottles, granulated sweetener, maple butter, t-shirts, recipe collections, candies, and coloring books for the kids. One table displays individually-packaged maple macarons almost like museum exhibits in cubes of clear plastic. To the left, a spectrum of maple bottles draws the eye like a stained glass window to the space beyond: a long, two-story section with vaulted rafters and polished concrete floors lined with barrels of distilled water. At the center of this open space, a gleaming steel evaporator over twenty feet long fills the room with sweet smells that ascend into the winter sky.

Nate Williams (Left) looks after equipment at Dutch Hill Maple

Tending the evaporator was Nate Williams, who builds school sports fields in the summer and switches to maple every winter. After he moved to Syracuse, his father’s annual ritual of boiling sap in the Finger Lakes was the magnet that kept him coming back. Nate and his wife Cristy started tapping trees on Dutch Hill in 2015 then purchased more land across the creek. They now tap 8,500 maples, using a system of sensors, pumps, and filters to convert roughly a hundred thousand gallons of sap (in a good year) into maple syrup.

Maple farming is a partnership of reciprocity between forests and families if they care for each other through the generations. Nate’s father, now in his seventies, still taps over a thousand trees himself. He revisits each tree at least twice a year to set up the taps then take them down again before bacteria can form in warmer weather.

Nate is halfway through the maple farmer’s litany of sugar-making when he says something I hadn’t heard in my month of visiting farms: “the trees give us something else we can’t get any other way: pure, distilled water.” With the farm so close to Tully’s brine fields, even a well that reached the depths of the hill would only yield a slurry of sulfur. So Dutch Hill Maple treasures a gift from the trees, a missing ingredient that other farmers steam or pour away: the water in the sap.

Pete Seeger’s classic song “Maple Syrup Time” is bio-chemically correct that tree sap is roughly 97% water. Maple trees use osmosis to draw water from rain and snow through their roots to distribute sugar to where it’s needed. If you have ever gulped energy drink on a hot day, you understand that our bodies use water to transport sugar. That’s why I fill my bottles with a mix of water, maple syrup, and electrolytes to fuel my long winter rides.

But maple technology has advanced since Seeger wrote the song. Farmers no longer boil all the water away. At Dutch Hill, Nate shows me the reverse osmosis filters that separate sugar from sap. At one end, the resulting almost-syrup flows into the evaporator for a final finishing boil. At the other end emerges pure, distilled water. Other farms in New York have started to market “tree-filtered water” as an alternative to spring water. At Dutch Hill Maple, this purified sap is the farm’s main source of water, on land with no viable aquifer.

On my ride home through a landscape glistening with fresh snow, I re-enter the realm of Maple Syrup Time and think of Seeger’s line “As in life or revolution, rarely is there a quick solution.” That’s been true for the Tully Valley too.

In the forty years since the mine closed, investigations and lawsuits have forced the company and the state to spend nearly a billion dollars cleaning Onondaga Lake downstream. The cleanup has reduced the lake’s level of mercury, among other chemistry in the sediments. In time, the fish may become safe to eat again. Just last year, governor Hochul announced that a thousand acres of Honeywell land will be granted back to the Onondaga Nation, who see the creek as sacred. Geologists report that the bedrock fractures above the brine fields have stabilized, though they can’t promise what the future holds. And the forest continues to grow.

If you ever need to shelter from the snow in the town of Tully south of Syracuse near the old Honeywell mine, park your bike at the porch of the Bloomin Cup Cafe and ask for a Toasted Maple latte. The baristas will proudly explain that the Dutch Hill Maple syrup is from forests nearby, with a level of adoration reserved only for treats that locals truly enjoy.

What does hope taste like to you? The Toasted Maple is a multi-layered wonder, with a thick froth from the smoky sweetness of fresh maple syrup. But the liquid I’ll most remember from my visit is a barrel of fresh water on a hill stitched together by the roots of maple trees, the gift of a forest that outlasted the mine that broke the land.

I am grateful to the geologists who provided expert advice and fact-checking on the chemistry and geology of the Tully Valley area. Thank you!

Firms spin scientific results as proof of safety while also undermining the generalizability of the findings. Where do we go from here?

Do a round of scientific studies show, as Meta claims, that Facebook algorithms didn’t cause much harm during the 2020 election? Or should we believe the scientists themselves, some of whom disagree?

This controversy became more complicated this week with news that the company was changing the algorithm underneath the researchers while it was happening — making it hard to draw any general claims.

Sadly, scientists could have predicted this conflict before the 2020 election project even got underway.

To understand why, you have to understand how recommender systems work. Recommender systems are assemblages of many models, which engineers/data scientists regularly update, as Arvind Naraynan writes in an excellent summary for the Knight First Amendment Institute.

This complexity makes it unlikely for a study to have a consistent “intervention” throughout. Randomized trials with recommender systems are unlikely to be internally consistent, since the algorithm keeps changing. In my 2016 work on influencing recommenders to reduce the spread of unreliable news, I had to restart a study when Reddit’s algorithm changed. I used this opportunity to write about the generalizability problem. While scientists on the 2020 Election project carefully noted this issue, Meta however has not been so honest about the results of the 2020 studies.

Changing algorithms create another problem: research findings may not apply for long, even if they are valid. That’s something that I along with Rob Kitchin, Kevin Munger, and others have been arguing for years, in my case across multiple field experiments and an article in Nature.

• Matias, J. N. (2023). Humans and algorithms work together — so study them together. Nature, 617(7960), 248–251.
• Munger, K. (2019). The limited value of non-replicable field experiments in contexts with low temporal validity. Social Media+ Society, 5(3), 2056305119859294.
• Kitchin, R. (2017). Thinking critically about and researching algorithms. Information, Communication, and Society.

Until we solve this problem of intervention validity in independent research about adaptive algorithms (like recommender systems), scientists won’t be able to provide reliable answers about the effects of a system that can be used to guide future decisions. And as we have seen in this week’s news, companies are not incentivized to produce that knowledge:

• their business models require them to change the algorithms regularly
• the algorithms adapt to changing user behavior and contexts
• firms have reputational and regulatory incentives to support strong but unreliable causal claims

The authors of the scientific studies in the 2020 Facebook Elections project are aware of this and wrote very clear qualifications to their findings. And the critics are right to point out this important issue. Rightly or wrongly, people expect randomized trials to provide gold-standard answers that stand the the test of time. That’s how Meta PR is trying to spin the results, and unfortunately, the state of the art in science does not make such strong claims possible at this time.

I see this as an important challenge for science, policy, and for tech leaders who are willing to think long-term about the health of the industry. Society needs causal knowledge, and we need to create the conditions for independent knowledge that people can trust, as I have been arguing for nearly a decade now.

The good news is that these valuable 2020 election studies do actually move science forward toward answers, even if they can’t provide those answers.

In the meantime, policymakers and the public are right to be concerned about this mismatch between what they expect from randomized trials and what they got from this project — in addition to their reasonable fears of corporate skullduggery, whatever the inside story is. Despite frustrating and painful disagreements, I hope this debate galvanizes people to collaborate and moves the conversation forward.

Image source: Stock Catalog

Earlier this year, my collaborator Jonathan Zong sent me a beautiful article by the theologian of technology Hanna Reichel about the ethics and design of blades, and what it teaches us about the design of ideas. Quoting the book of Isaiah chapter 2, she reminds us that swords can be changed to plowshares — and that the reverse is also true:

“Imagine you found a piece of metal with a sharp edge. You turn it in your hands. What you imagine that you could do with it will depend…”

What happens next in this story? Will the edge be used to threaten another person, mark a path, or release nutrients into the soil with a new year? Reichel argues that critical scholars can “read tradition against tradition” to imagine transformations of this generative kind.

The idea of turning over ideas on top of themselves with a hoe (or a plow) is a beautiful metaphor for critical thinking and critical design. As a professor, this is one of my jobs — to nurture the creative, critical imaginations of those I mentor, supporting them to imagine and develop new growth.

[…read more at the Citizens and Technology Lab…]