Robin Smith, Author at Research Blog

Sure, A.I. chatbots can write emails, summarize an article, or come up with a grocery list. But ChatGPT-style artificial intelligence and other machine learning techniques have been making their way into another realm: healthcare.

Imagine using AI to detect early changes in our health before we get sick, or understand what happens in our brains when we feel anxious or depressed — even design new ways to fight hard-to-treat diseases.

These were just a few of the research themes discussed at the Duke Summit on AI for Health Innovation, held October 9 – 11.

Duke assistant professor Pranam Chatterjee is the co-founder of Gameto, Inc. and UbiquiTx, Inc. Credit: Brian Strickland

For assistant professor of biomedical engineering Pranam Chatterjee, the real opportunity for the large language models behind tools like ChatGPT lies not in the language of words, but in the language of biology.

Just like ChatGPT predicts the order of words in a sentence, the language models his lab works on can generate strings of molecules that make up proteins.

His team has trained language models to design new proteins that could one day fight diseases such as Huntington’s or cancer, even grow human eggs from stem cells to help people struggling with infertility.

“We don’t just make any proteins,” Chatterjee said. “We make proteins that can edit any DNA sequence, or proteins that can modify other disease-causing proteins, as well as proteins that can make new cells and tissues from scratch.”

Duke assistant professor Monica Agrawal is the co-founder of Layer Health. Credit: Brian Strickland

New faculty member Monica Agrawal said algorithms that leverage the power of large language models could help with another healthcare challenge: mining the ever-expanding trove of data in a patient’s medical chart.

To choose the best medication for a certain patient, for example, a doctor might first need to know things like: How has their disease progressed over time? What interventions have already been tried? What symptoms and side effects did they have? Do they have other conditions that need to be considered?

“The challenge is, most of these variables are not found cleanly in the electronic health record,” said Agrawal, who joined the departments of computer science and biostatistics and bioinformatics this fall.

Instead, most of the data that could answer these questions is trapped in doctors’ notes. The observations doctors type into a patient’s electronic medical record during a visit, they’re often chock-full of jargon and abbreviations.

The shorthand saves time during patient visits, but it can also lead to confusion among patients and other providers. What’s more, reviewing these records to understand a patient’s healthcare history is time-intensive and costly.

Agrawal is building algorithms that could make these records easier to maintain and analyze, with help from AI.

“Language is really embedded across medicine, from notes to literature to patient communications to trials,” Agrawal said. “And it affects many stakeholders, from clinicians to researchers to patients. The goal of my new lab is to make clinical language work for everyone.”

Duke assistant professor Jessilyn Dunn leads Duke’s BIG IDEAs Lab. Credit: Brian Strickland

Jessilyn Dunn, an assistant professor of biomedical engineering and biostatistics and bioinformatics at Duke, is looking at whether data from smartwatches and other wearable devices could help detect early signs of illness or infection before people start to have symptoms and realize they’re sick.

Using AI and machine learning to analyze data from these devices, she and her team at Duke’s Big Ideas Lab say their research could help people who are at risk of developing diabetes take action to reverse it, or even detect when someone is likely to have RSV, COVID-19 or the flu before they have a chance to spread the infection.

“The benefit of wearables is that we can gather information about a person’s health over time, continuously and at a very low cost,” Dunn said. “Ultimately, the goal is to provide patient empowerment, precision therapies, just-in-time intervention and improve access to care.”

Duke associate professor David Carlson. Credit: Brian Strickland

David Carlson, an associate professor of civil and environmental engineering and biostatistics and bioinformatics, is developing AI techniques that can make sense of brain wave data to better understand different emotions and behaviors.

Using machine learning to analyze the electrical activity of different brain regions in mice, he and his colleagues have been able to track how aggressive a mouse is feeling, and even block the aggression signals to make them more friendly to other mice.

“This might sound like science fiction,” Carlson said. But Carlson said the work will help researchers better understand what happens in the brains of people who struggle with social situations, such as those with autism or social anxiety disorder, and could even lead to new ways to manage and treat psychiatric disorders such as anxiety and depression.

Can You Spot the Species in These Lemur Lookalikes?

By Robin Smith

On October 7, 2024

In Animals, Genetics/Genomics, Lemurs

In some parts of the world, animals are going extinct before scientists can even name them.

Such may be the case for mouse lemurs, the saucer-eyed, teacup-sized primates native to the African island of Madagascar.

Various species of mouse lemurs found in Madagascar. Photos by Sam Hyde Roberts

There, deforestation has prompted the International Union for the Conservation of Nature (IUCN) to classify some of these tree-dwelling cousins as “endangered” even before they are formally described.

Duke professor Anne Yoder has been trying to take stock of how many mouse lemur species are alive today before they blink out of existence.

It’s not an easy task. Mouse lemurs are shy, they only come out at night, and they live in hard-to-reach places in remote forests. To add to the difficulty, many species of mouse lemurs are essentially lookalikes. It’s impossible to tell them apart just by peering at them through binoculars.

When Yoder first started studying mouse lemurs some 25 years ago, there were only three distinct species recognized by scientists. Over time and with advances in DNA sequencing, researchers began to wonder if what looked like three species might actually be upwards of two dozen.

In a new study, Yoder and dozens of colleagues from Europe, Madagascar and North America compiled and analyzed 50 years of hard-won data on the physical, behavioral and genetic differences among mouse lemurs to try to pin down the true number.

While many mouse lemur species look alike, they have different diets, and males use different calls to find and woo their mates, the researchers explain.

By pinning down their number and location, researchers hope to make more informed decisions about how best to help keep these species from the brink.

The study was published Sept. 27 in the journal Nature Ecology & Evolution.

Braxton Craven Distinguished Professor of Evolutionary Biology, Anne Yoder was director of the Duke Lemur Center from 2006 to 2018.

Student Wealth and Poverty Across Durham Public Schools, Mapped

By Robin Smith

On August 20, 2024

In Data, Students, Visualization

New maps of Durham released by students in Duke’s Data+ research program show the Bull City as a patchwork of red, white and pink. But what looks like a haphazardly assembled quilt is actually a picture of the socioeconomic realities facing Durham’s 32,000-plus public school students.

The color patches represent the home values across Durham, showing roughly where more and less affluent students live. The darker the red, the higher-priced their housing.

Like cities and neighborhoods, schools face economic disparities too. Research shows that school segregation by race and class in North Carolina has gotten steadily worse over the last three decades.

A 2024 study by North Carolina State University revealed that the typical low-income student attends schools where more than 70% of their classmates are low-income too — a trend that worsens the achievement gap between the richest and poorest children.

A new student assignment plan that Durham Public Schools is rolling out this year aims to combat that trend by redrawing district boundary lines — the thick black lines on the map — to make schools more diverse and equitable.

But if schools are to tackle economic segregation, they’ll need accurate ways to measure it as Durham continues to grow and change.

As kids across Durham head back to class, some 1500 elementary school students are changing schools this year under the Durham Public Schools Growing Together plan, which aims to increase equity and access across schools.

That was the challenge facing a team in Duke’s Data+ program this summer. For 10 weeks, Duke students Alex Barroso and Dhaval Potdar collaborated with school planners at Durham Public Schools to look at how family wealth and poverty are distributed across the school system.

“Socioeconomic status is a complicated thing,” said Barroso, a Duke junior majoring in statistical science.

For years, the standard way to identify children in need was using free and reduced-price lunch statistics from the National School Lunch Program, along with published income data from the U.S. Census Bureau.

But those numbers can be unreliable, Barroso said.

Changing state and federal policies mean that more districts — including Durham Public Schools — are providing free meals to all students, regardless of their family income. But as a result, schools no longer have an exact count of how many students qualify.

And Census estimates are based on geographic boundaries that can mask important variation in the data when we look more closely.

At a symposium in Gross Hall in July, Barroso pointed to several dark red patches (i.e., more expensive housing) bordering white ones (i.e., more affordable) on one of the team’s maps.

In some parts of the city, homes worth upwards of a million dollars abut modest apartments worth a fraction of that, “which can skew the data,” he said.

The problem with Census estimates “is that everyone who lives in that area is reported as having the same average income,” said team lead Vitaly Radsky, a PhD student at UNC’s School of Education and school planner with Durham Public Schools.

So they took a different approach: using homes as a proxy for socioeconomic status.

Research has confirmed that students from higher-value homes perform better in school as measured by standardized math tests.

The team created a custom script that fetches publicly available data on every home in Durham from sources such as Durham Open Data and the Census, and then automatically exports it to a dashboard that shows the data on a map.

“Every single house is accounted for within this project,” Barroso said.

They ran into challenges. For example, Census data are tied to tracts that don’t necessarily align with the district boundaries used by schools, said Dhaval Potdar, a graduate student in Duke’s Master in Interdisciplinary Data Science.

One takeaway from their analysis, Potdar said, is no one yardstick sums up the economic well-being of every student.

In Durham, the typical public school student lives in a home valued at about $300,000.

But the picture varies widely when you zoom in on different geographic scales and footprints.

It’s also a different story if you account for the significant fraction of Durham families who live among neighbors in a larger building such as an apartment, townhouse or condominium, instead of a single-family home.

Considering a home’s age can change the picture too.

Generally speaking, students who live in more expensive homes come from more affluent families. But in many parts of the U.S., home prices have far outpaced paychecks. That means a home that has soared in value in the years since it was purchased may not reflect a family’s true economic situation today, particularly if their income remained flat.

The team’s data visualizations aim to let school planners look at all those factors.

There are still issues to be ironed out. For example, there’s some work to be done before planners can make apples-to-apples comparisons between a student whose family owns their home versus renting a similar property, Barroso said.

“No data source is perfect,” but the research offers another way of anticipating the shifting needs of Durham students, Radsky said.

“The traditional metrics really aren’t getting at the granular fabric of the Durham community,” said Mathew Palmer, the district’s senior executive director of school planning and operational services.

Research like this helps address questions like, “are we putting our resources where the kids need them the most? And are schools equitable?”

“This analysis gives schools more tools moving forward,” Palmer said.

A Camera Trap for the Invisible

By Robin Smith

On July 29, 2024

In Artificial Intelligence, Physics

It sounds fantastical, but it’s a reality for the scientists who work at the world’s largest particle collider:

In an underground tunnel some 350 feet beneath the France–Switzerland border, a huge device called the Large Hadron Collider sends beams of protons smashing into each other at nearly the speed of light, creating tiny eruptions that mimic the conditions that existed immediately after the Big Bang.

Scientists like Duke physicist Ashutosh Kotwal think the subatomic debris of these collisions could contain hints of the universe’s “missing matter.” And with some help from artificial intelligence, Kotwal hopes to catch these fleeting clues on camera.

A view inside the ATLAS detector at the Large Hadron Collider. Akin to a giant digital camera, physicists hope to use the detector in the quest to find dark matter, the mysterious stuff that fills the universe but no one has ever seen it. Credit: CERN.

Ordinary matter — the stuff of people and planets — is only part of what’s out there. Kotwal and others are hunting for dark matter, an invisible matter that’s five times more abundant than the stuff we can see but whose nature remains a mystery.

Scientists know it exists from its gravitational influence on stars and galaxies, but other than that we don’t know much about it.

The Large Hadron Collider could change that. There, researchers are looking for dark matter and other mysteries using detectors that act like giant 3D digital cameras, taking continuous snapshots of the spray of particles produced by each proton-proton collision.

Only ordinary particles trigger a detector’s sensors. If researchers can make dark matter at the LHC, scientists think one way it could be noticeable is as a sort of disappearing act: heavy charged particles that travel a certain distance — 10 inches or so — from the point of collision and then decay invisibly into dark matter particles without leaving a trace.

If you retraced the paths of these particles, they would leave a telltale “disappearing track” that vanishes partway through the detector’s inner layers.

When beams collide at the Large Hadron Collider, they split into thousands of smaller particles that fly out in all directions before vanishing. Scientists think some of those particles could make up dark matter, and Duke physicist Ashutosh Kotwal is using AI and image recognition to help in the hunt. Credit: Pcharito.

But to spot these elusive tracks they’ll need to act fast, Kotwal says.

That’s because the LHC’s detectors take some 40 million snapshots of flying particles every second.

That’s too much raw data to hang on to everything and most of it isn’t very interesting. Kotwal is looking for a needle in a haystack.

“Most of these images don’t have the special signatures we’re looking for,” Kotwal said. “Maybe one in a million is one that we want to save.”

Researchers have just a few millionths of a second to determine if a particular collision is of interest and store it for later analysis.

“To do that in real time, and for months on end, would require an image recognition technique that can run at least 100 times faster than anything particle physicists have ever been able to do,” Kotwal said.

Kotwal thinks he may have a solution. He has been developing something called a “track trigger,” a fast algorithm that is able to spot and flag these fleeting tracks before the next collision occurs, and from among a cloud of tens of thousands of other data points measured at the same time.

Ashutosh Kotwal is the Fritz London Distinguished Professor of Physics at Duke University.

His design works by divvying up the task of analyzing each image among a large number of AI engines running simultaneously, built directly onto a silicon chip. The method processes an image in less than 250 nanoseconds, automatically weeding out the uninteresting ones.

Kotwal first described the approach in a sequence of two papers published in 2020 and 2021. In a more recent paper published this May in Scientific Reports, he and a team of undergraduate student co-authors show that his algorithm can run on a silicon chip.

Kotwal and his students plan to build a prototype of their device by next summer, though it will be another three or four years before the full device — which will consist of about 2000 chips — can be installed at detectors at the LHC.

As the performance of the accelerator continues to crank up, it will produce even more particles. And Kotwal’s device could help make sure that, if dark matter is hiding among them, scientists won’t miss it.

“Our job is to ensure that if dark matter production is happening, then our technology is up to snuff to catch it in the act,” Kotwal said.

Duke Mathletes Stand Out in a Crowd

By Robin Smith

On July 12, 2024

In Mathematics, Students

Standing out in a crowd of competitors is no easy task. But one Duke team has done just that — in math.

The Blue Devils were the only U.S.-based team to claim a top 25 finish at the 40th annual Mathematical Contest in Modeling (MCM), beating out more than 18,500 other teams from 20 countries.

*Blue Devils Brandon Lu, Benny Sun and Chris Kan (L to R) finished in the top 25 out of 18,525 teams in an international math contest called Mathematical Contest in Modeling.*

The team consisted of undergraduates Christopher Kan, Benny Sun, and Brandon Lu. Their task: to solve a real-world problem using mathematical modeling within 96 hours.

This year’s contestants tackled problems ranging from analyzing what gives tennis players an edge at Wimbledon, to optimizing search and rescue operations for missing submersibles.

The Duke team tackled a challenge that has vexed the fishing industry in the Great Lakes: predicting the impact of an invasive parasitic fish called the sea lamprey that can wreak havoc on native fish.

By adapting existing models from biology and biochemistry to model the sea lamprey population, the students were able to determine how to best apply treatments to rid streams of these parasites.

*Two sea lampreys chewing their way into the flesh of a native lake trout of the Great Lakes. (Great Lakes Fisheries Commission)*

The contest “is much more open-ended and creatively-focused than most STEM classes,” said Sun, a mathematics and computer science double major at Duke.

The participants try out different approaches to modeling the problems, and there is no one correct answer.

Sun, Kan and Lu also received the Mathematical Association of America Award for their paper. “They did a great job,” said team advisor Veronica Ciocanel, an assistant professor of math and biology who also co-organizes a local version of the contest each fall, called the Triangle Competition in Math Modeling.

In these contests, creativity, time management and writing skills are just as important as cramming on concepts.

“We realized that communication was as important as the findings themselves,” Sun said. “We spent the last two days primarily focused on writing a good paper.”

Having fun as a team is important too, Sun said. “Team chemistry can be an especially important factor in success when you are all locked in the same room for the weekend.”

Wiring the Brain

By Robin Smith

On June 26, 2024

In Animals, Neuroscience, Students

From tiny flies, Duke researchers are finding new clues to how the brain sets up its circuitry.

In her time at Duke, Khanh Vien figures she’s dissected close to 10,000 fly brains. For her PhD she spent up to eight hours each day peering at baby flies under the microscope, teasing out tiny brains a fraction the size of a poppy seed.

“I find it very meditative,” she said.

Vien acknowledges that, to most people, fruit flies are little more than a kitchen nuisance; something to swat away. But to researchers like her, they hold clues to how animal brains — including our own — are built.

While the human brain has some 86 billion neurons, a baby fruit fly’s brain has a mere 3016 — making it millions of times simpler. Those neurons talk to each other via long wire-like extensions, called axons, that relay electrical and chemical signals from one cell to the next.

Vien and other researchers in Professor Pelin Volkan’s lab at Duke are interested in how that wiring gets established during the fly’s development.

By analyzing a subset of neurons responsible for the fly’s sense of smell, the researchers have identified a protein that helps ensure that new neurons extend their axons to the correct spots in the olfactory area of the young fly’s brain and not elsewhere.

Because the same protein is found across the animal kingdom, including humans, the researchers say the work could ultimately shed light on what goes awry within the brains of people living with schizophrenia and other mental illnesses.

Their findings are published in the journal iScience.

Khanh Vien earned her PhD in developmental and stem cell biology in Professor Pelin Volkan’s lab at Duke.

To get a fuller picture of a forest, sometimes research requires a team effort

By Robin Smith

On April 1, 2024

In Animals, Biology, Climate/Global Change, Environment/Sustainability, Field Research, Lemurs

Film by Riccardo Morrelas, Zahava Production

For some people, the word “rainforest” conjures up vague notions of teeming jungles. But Camille DeSisto sees something more specific: a complex interdependent web.

For the past few years, the Duke graduate student has been part of a community-driven study exploring the relationships between people, plants and lemurs in a rainforest in northern Madagascar, where the health of one species depends on the health of others.

Many lemurs, for example, eat the fruits of forest trees and deposit their seeds far and wide in their droppings, thus helping the plants spread. People, in turn, depend on the plants for things like food, shelter and medicines.

But increasingly, deforestation and other disturbances are throwing these interactions out of whack.

DeSisto and her colleagues have been working in a 750,000-acre forest corridor in northeast Madagascar known as the COMATSA that connects two national parks.

The area supports over 200 tree species and nine species of lemurs, and is home to numerous communities of people.

A red-bellied lemur (*Eulemur rubriventer*) in a rainforest in northeast Madagascar. Photo by Martin Braun.

“People live together with nature in this landscape,” said DeSisto, who is working toward her Ph.D. in ecology at the Nicholas School of the Environment.

But logging, hunting and other stressors such as poverty and food insecurity have taken their toll.

Over the last quarter century, the area has lost 14% of its forests, mostly to make way for vanilla and rice.

This loss of wild habitats risks setting off a series of changes. Fewer trees also means fewer fruit-eating lemurs, which could create a feedback loop in which the trees that remain have fewer opportunities to replace themselves and sprout up elsewhere — a critical ability if trees are going to track climate change.

DeSisto and her colleagues are trying to better understand this web of connections as part of a larger effort to maximize forest resilience into an uncertain future.

To do this work, she relies on a network of a different sort.

The research requires dozens of students and researchers from universities in Madagascar and the U.S., not to mention local botanists and lemur experts, the local forest management association, and consultants and guides from nearby national parks, all working together across time zones, cultures and languages.

Forest field team members at camp (not everyone present). Photo credit: Jane Slentz-Kesler.

Together, they’ve found that scientific approaches such as fecal sampling or transect surveys can only identify so much of nature’s interconnected web.

Many lemurs are small, and only active at night or during certain times of year, which can make them hard to spot — especially for researchers who may only be on the ground for a limited time.

To fill the gaps, they’re also conducting interviews with local community members who have accumulated knowledge from a lifetime of living on the land, such as which lemurs like to munch on certain plants, what parts they prefer, and whether people rely on them for food or other uses.

By integrating different kinds of skills and expertise, the team has been able to map hidden connections between species that more traditional scientific methods miss.

For example, learning from the expertise of local community members helped them understand that forest patches that are regenerating after clear-cutting attract nocturnal lemurs that may — depending on which fruits they like to eat — promote the forest’s regrowth.

Camille DeSisto after a successful morning collecting lemur fecal samples.

Research collaborations aren’t unusual in science. But DeSisto says that building collaborations with colleagues more than 9,000 miles away from where she lives poses unique challenges.

Just getting to her field site involves four flights, several bumpy car rides, climbing steep trails and crossing slippery logs.

“Language barriers are definitely a challenge too,” DeSisto said.

She’s been studying Malagasy for seven years, but the language’s 18 dialects can make it hard to follow every joke her colleagues tell around the campfire.

To keep her language skills sharp she goes to weekly tutoring sessions when she’s back in the U.S., and she even helped start the first formal class on the language for Duke students.

“I like to think of it as language opportunities, not just language barriers,” DeSisto said.”

“Certain topics I can talk about with much more ease than others,” she added. “But I think making efforts to learn the language is really important.”

When they can’t have face-to-face meetings the team checks in remotely, using videoconferencing and instant messaging to agree on each step of the research pipeline, from coming up with goals and questions and collecting data to publishing their findings.

“That’s hard to navigate when we’re so far away,” DeSisto said. But, she adds, the teamwork and knowledge sharing make it worth it. “It’s the best part of research.”

This research was supported by Duke Bass Connections (“Biocultural Sustainability in Madagascar,” co-led by James Herrera), Duke Global, The Explorers Club, Primate Conservation, Inc., Phipps Conservatory and Botanical Gardens, and the Garden Club of America.

A Grueling Math Test So Hard, Almost No One Gets a Perfect Score

By Robin Smith

On March 1, 2024

In Engineering, Mathematics, Students

Yet hundreds of schools compete each year, and this time the Blue Devils made it into the top three

Every year, thousands of college students from across the U.S. and Canada give up a full Saturday before finals begin to take a notoriously difficult, 6-hour math test — and not for a grade, but for fun.

In “the Putnam,” as it’s known, contestants spend two 3-hour sessions trying to solve 12 proof-based math problems worth 10 points apiece.

More than 150,000 people have taken the exam in the contest’s 85-year history, but only five times has someone earned a perfect score. Total scores of 1 or 0 are not uncommon.

Despite the odds, the Blue Devils had a strong showing this year.

A total of 3,857 students from 471 schools competed in the December contest. In results announced Feb. 16, a Duke team consisting of Erick Jiang ’26, James “Fletch” Rydell ’26 and Kaixin “Kai” Wang ’27 ranked third in North America behind MIT and Harvard, winning a $15,000 prize for Duke and each taking home $600 for themselves.

According to mathematics professor Lenny Ng, it’s Duke’s best performance in almost 20 years.

“This is the first time a Duke team has placed this high since 2005,” said Ng, who was a three-time Putnam Fellow himself, finishing in the top five each year he was an undergraduate at Harvard.

There’s no official syllabus for prepping for the Putnam. To get ready, the students practice working through problems and discussing their solutions in a weekly problem-solving seminar held each fall.

Students serve as the instructors, focusing on a different topic each week ranging from calculus to number theory.

“They get a sense of what the problems are like, so it’s not quite as intimidating as it might be if they went into the contest cold,” said math department chair Robert Bryant.

“Not only do they learn how to do the problems, but they also get to know each other,” said professor emeritus David Kraines, who has coached Duke Putnam participants for more than 30 years.

Kraines said 8-10 students take his problem-solving seminar for credit each fall. “We always get another 10 or so who come for the pizza,” Kraines said.

The biggest difference between a Putnam problem and a homework problem, said engineering student Rydell, is that usually with a homework problem you’ve already been shown what to do; you just have to apply it.

Whereas most of the time in math competitions like the Putnam, “there’s no clear path forward when you first see the problem,” Rydell said. “They’re more about finding some insight or way of looking at the problem in a different perspective.”

Putnam problems are meant to be solvable using only paper and pencil — no computing power required. The contestants work through each problem by hand, trying different paths towards a solution and spelling out their reasoning step-by-step.

This year, one problem involved determining how many configurations of coins are possible given a grid with coins sitting in some of the squares, if those coins are only allowed to move in certain ways.

Another question required knowing something about the geometry of a 20-sided shape known as a icosahedron.

“That was the one I struggled with the most,” said Wang, whose individual score nevertheless tied him for sixth place overall out of 3,857 contestants.

The most common question he gets asked about the Putnam, Rydell said, is not so much what’s on the test, but why people take it in the first place.

This year’s test was so challenging that a score of 78 out of 120 or better — just 65% — was enough to earn a spot in the top 10.

Most of the people who took it scored less than 10%, which means many problems went unsolved.

“For days after I took the Putnam, I would think about the problems and wonder: could I have done it better this way? You can become obsessed,” said Bryant, who took the Putnam in the 1970s as a college student at NC State.

Sophomores Jiang and Rydell, who both ranked in the top 5%, see it as an opportunity to “meet people who also enjoy problem solving,” Jiang said.

“I’m not a math major so I probably wouldn’t do much of this kind of problem solving otherwise,” Rydell said.

For Rydell it’s also the aha moment: “Just the reward of when you solve a problem, the feeling of making that breakthrough,” Rydell said.

Professor Kraines’ weekly problem-solving seminar, MATH 283S, takes place on Tuesday evenings at 6:15 p.m. during the fall semester. Registration for Fall 2024 begins April 3.

Putting Stronger Guardrails Around AI

By Robin Smith

On November 21, 2023

In Artificial Intelligence, Computers/Technology, Engineering, Law, Lecture, Policy

AI regulation is ramping up worldwide. Duke AI law and policy expert Lee Tiedrich discusses where we’ve been and where we’re going.

DURHAM, N.C. — It’s been a busy season for AI policy.

The rise of ChatGPT unleashed a frenzy of headlines around the promise and perils of artificial intelligence, and raised concerns about how AI could impact society without more rules in place.

Consequently, government intervention entered a new phase in recent weeks as well. On Oct. 30, the White House issued a sweeping executive order regulating artificial intelligence.

The order aims to establish new standards for AI safety and security, protect privacy and equity, stand up for workers and consumers, and promote innovation and competition. It’s the U.S. government’s strongest move yet to contain the risks of AI while maximizing the benefits.

“It’s a very bold, ambitious executive order,” said Duke executive-in-residence Lee Tiedrich, J.D., who is an expert in AI law and policy.

Tiedrich has been meeting with students to unpack these and other developments.

“The technology has advanced so much faster than the law,” Tiedrich told a packed room in Gross Hall at a Nov. 15 event hosted by Duke Science & Society.

“I don’t think it’s quite caught up, but in the last few weeks we’ve taken some major leaps and bounds forward.”

Countries around the world have been racing to establish their own guidelines, she explained.

The same day as the US-led AI pledge, leaders from the Group of Seven (G7) — which includes Canada, France, Germany, Italy, Japan, the United Kingdom and the United States — announced that they had reached agreement on a set of guiding principles on AI and a voluntary code of conduct for companies.

Both actions came just days before the first ever global summit on the risks associated with AI, held at Bletchley Park in the U.K., during which 28 countries including the U.S. and China pledged to cooperate on AI safety.

“It wasn’t a coincidence that all this happened at the same time,” Tiedrich said. “I’ve been practicing law in this area for over 30 years, and I have never seen things come out so fast and furiously.”

The stakes for people’s lives are high. AI algorithms do more than just determine what ads and movie recommendations we see. They help diagnose cancer, approve home loans, and recommend jail sentences. They filter job candidates and help determine who gets organ transplants.

Which is partly why we’re now seeing a shift in the U.S. from what has been a more hands-off approach to “Big Tech,” Tiedrich said.

Tiedrich presented Nov. 15 at an event hosted by Duke Science & Society.

In the 1990s when the internet went public, and again when social media started in the early 2000s, “many governments — the U.S. included — took a light touch to regulation,” Tiedrich said.

But this moment is different, she added.

“Now, governments around the world are looking at the potential risks with AI and saying, ‘We don’t want to do that again. We are going to have a seat at the table in developing the standards.’”

Power of the Purse

Biden’s AI executive order differs from laws enacted by Congress, Tiedrich acknowledged in a Nov. 3 meeting with students in Pratt’s Master of Engineering in AI program.

Congress continues to consider various AI legislative proposals, such as the recently introduced bipartisan Artificial Intelligence Research, Innovation and Accountability Act, “which creates a little more hope for Congress,” Tiedrich said.

What gives the administration’s executive order more force is that “the government is one of the big purchasers of technology,” Tiedrich said.

“They exercise the power of the purse, because any company that is contracting with the government is going to have to comply with those standards.”

“It will have a trickle-down effect throughout the supply chain,” Tiedrich said.

The other thing to keep in mind is “technology doesn’t stop at borders,” she added.

“Most tech companies aren’t limiting their market to one or two particular jurisdictions.”

“So even if the U.S. were to have a complete change of heart in 2024” and the next administration were to reverse the order, “a lot of this is getting traction internationally,” she said.

“If you’re a U.S. company, but you are providing services to people who live in Europe, you’re still subject to those laws and regulations.”

From Principles to Practice

Tiedrich said a lot of what’s happening today in terms of AI regulation can be traced back to a set of guidelines issued in 2019 by the Organization for Economic Cooperation and Development, where she serves as an AI expert.

These include commitments to transparency, inclusive growth, fairness, explainability and accountability.

For example, “we don’t want AI discriminating against people,” Tiedrich said. “And if somebody’s dealing with a bot, they ought to know that. Or if AI is involved in making a decision that adversely affects somebody, say if I’m denied a loan, I need to understand why and have an opportunity to appeal.”

“The OECD AI principles really are the North Star for many countries in terms of how they develop law,” Tiedrich said.

“The next step is figuring out how to get from principles to practice.”

“The executive order was a big step forward in terms of U.S. policy,” Tiedrich said. “But it’s really just the beginning. There’s a lot of work to be done.”

New Blogger Noor Nazir: Mental Health With a Pakistani Twist

By Robin Smith

On September 28, 2023

In Behavior/Psychology, Economics, Students

My name is Door but replace the ‘D’ with an ‘N’.

Yes, I’m Noor and yes again, that is exactly how I introduced my freshman self to everyone in my year. Before you wonder, it’s an Arabic name and no I’m not from the Middle East! I’m a die-hard Pakistani with an overwhelming – and embarrassing – amount of love for Taylor Swift and Local Pakistani Music (stream Talha Anjum, you’ll be surprised!).

My personality mainly encompasses my thirteen-month-old niece, Alaya. I like to think she’s my mini doppelganger (she is not) and the last eight months of my life have been encapsulated by her cute presence, smelly diapers and charming smile. We spend most of our time listening to Taylor Swift, and – sometimes – the nursery rhyme, One Little Finger. Other times, we play the guitar and sing for fun (your average Duke freshman).

Although, contrary to the ‘average Duke freshman’ who is sure about the trajectory of their next twenty years, I am not – at all. I find my mind wandering to several distinct fields of interest; whenever a classmate asks me “but where is your mind really at?”, my deliberate and circumspect answer is always “four to be exact: economics, political science, psychology and public policy”’. This answer is invariably met by an overt facial expression screaming their internal thought “oh so she’s really not sure”. But that side eye is beside the point since that uncertainty is precisely what led me to the Duke Research Blog.

In high school, whether it was the debate club or my interest in mental health, I always found a research angle to it. For debate, I’d research different case studies in order to formulate argumentation and rebuttals; for mental health, I’d utilize such case studies and would recreate what worked. My proudest creation, the Safe Space Society (a society in my alma matter, International School Lahore), was nothing short of a camaraderie and a community fostered with love and empathy. In my eyes, such a creation was only made possible because of extensive and life-long research by dedicated professionals.

Not only is research the perfect way to navigate my interests in a fulfilling manner, but it also acts as the tunnel vision to a transfigured world. Since my navigation wishes to find its destination in a declared major, I’m incredibly excited to write and learn about research revolving science, mental health, and anything Duke brings my way.

I am, however, most excited to translate and decode complex and seemingly mundane ideas in a nuanced and amusing way. The blog seems to be on a mission to make potential engineers excited about the next big thing in mental health research; this is a mission I’m excited and honored to take part in. To sum it up, my goal at Duke Research Blog is to attend the research events you don’t want to and then write about them to make you regret not attending those events!

– A serious warning: you will see me bringing a Pakistani twist to every article I write! It’s just what us Pakistanis do (for a sample look at the sentence above). –

Following the people and events that make up the research community at Duke

Author: Robin Smith Page 1 of 8

Duke experts discuss the potential of AI to help prevent, detect and treat disease

Can You Spot the Species in These Lemur Lookalikes?

Student Wealth and Poverty Across Durham Public Schools, Mapped

A Camera Trap for the Invisible

Duke Mathletes Stand Out in a Crowd

Wiring the Brain

To get a fuller picture of a forest, sometimes research requires a team effort

A Grueling Math Test So Hard, Almost No One Gets a Perfect Score

Putting Stronger Guardrails Around AI

New Blogger Noor Nazir: Mental Health With a Pakistani Twist