Patchen Barss

Groundbreakers: U of T initiative brings together experts to address major societal issues

Christopher.Sorensen — Thu, 02 Sep 2021 19:45:14 +0000

Groundbreakers: U of T initiative brings together experts to address major societal issues

Christopher.Sorensen Thu, 09/02/2021 - 15:45

Cutline

An expert in assistive robotics, Goldie Nejat works with a diverse group of researchers at U of T's Robotics Institute, one of several Institutional Strategic Initiatives, to build machines that can help seniors (photo courtesy of Goldie Nejat)

Patchen Barss

Topic

Our Community

Groundbreakers

Institutional Strategic Initiatives

Temerty Faculty of Medicine

Black

Faculty of Applied Science & Engineering

Molecular Genetics

Research & Innovation

Robotics

U of T Mississauga

U of T Scarborough

Women and Gender Studies

Goldie Nejat envisions a world where growing old is improved by robots that care.

The Canada Research Chair in Robots for Society and professor of mechanical engineering in the �ؿ�ζSM’s Faculty of Applied Science & Engineering is an expert in assistive robotics – machines that provide care, interventions and even companionship for long-term care residents and people living with dementia and other cognitive and physical impairments.

The work is extraordinarily complex. In addition to engineering and artificial intelligence, creating robotic companions for an aging population often requires expertise in gerontology, cognitive neuroscience and medicine – even ethics.

Enter the Robotics Institute. Part of U of T’s Institutional Strategic Initiatives (ISI) program, the institute brings together top experts in different fields – from hardware design to public policy – to solve thorny, robotics-related problems in applications ranging from health care and transportation to manufacturing and logistics.

“Being at U of T, you have access to world-class researchers,” says Nejat. “Every time I want to find a collaborator with specific expertise and knowledge, it's really easy.

“A lot of us [robotics engineers] work with surgeons, doctors, physiotherapists, occupational therapists – you name it.”

The Robotics Institute is just one of nearly two dozen institutional strategic initiatives launched by U of T to address complex, real-world challenges that cut across fields of expertise. Each initiative brings together a flexible, multidisciplinary team of researchers, students and partners from industry, government and the community to take on a “grand challenge.”

“The goal is to bring scientists, researchers and the broader research community together to achieve something that no one scientist or scientists from a single discipline could achieve on their own,” says Christine Allen, U of T’s associate vice-president and vice-provost, strategic initiatives.

“Whether they’re designing a new cancer treatment, developing more sustainable building materials, identifying better education systems for students with mental health issues, ISI participants are always exploring who can contribute – and who can benefit.

“Cross-divisional research and collaboration is paramount to addressing complex questions and making an urgently needed transformational impact.”

“A lot of us [robotics engineers] work with surgeons, doctors, physiotherapists, occupational therapists – you name it,” Nejat says (photo by �ؿ�ζSM)

The approach marks a departure from the traditional university model of loosely affiliated, but independent – and relatively siloed – faculties, departments, institutes and colleges. The problem, Allen says, is that modern societal challenges don’t map very well onto traditional university structures. Climate change, mental health, systemic racism, artificial intelligence and many other issues are so multifaceted and interconnected that they can only be addressed with diverse expertise, insights, and approaches.

The ISI approach, by contrast, seeks to take advantage of U of T’s size and breadth of expertise by creating research networks that cut across campuses and disciplines as varied as sociology and computer science – all in an effort to unlock answers to challenging problems and create new knowledge.

The Black Research Network, for instance, uses ISI’s multidisciplinary approach to help redress the sense of isolation that many Black faculty members feel by connecting them not only to each other, but also to the broader community – and resources – at the university.

“My work is to make sure the right conversations are happening, that the right clusters are coming together to produce small and large-scale research grants and teams,” says Beth Coleman, an associate professor of data and cities at U of T Mississauga’s Institute of Communication, Culture, Information and Technology and the Faculty of Information, and the network’s director. “In STEM fields especially, the number of Black professors is small and people tend to be really isolated.

“That isolation is part of why people succeed or don’t succeed. Some of that has to do with how comfortable you feel and how invited you feel.”

The co-founders of the network come from across the university: Rhonda McEwen, dean of U of T Mississauga; Maydianne Andrade, a professor of evolutionary biology and U of T Scarborough’s vice-dean of faculty affairs, equity and success; Lisa Robinson, associate dean, inclusion and diversity in the Temerty Faculty of Medicine, and Alissa Trotz, director of the Women & Gender Studies Institute.

The co-founders of the Black Research Network are (clockwise from top left): Professors Maydianne Andrade, Lisa Robinson, Alissa Trotz and Rhonda McEwen (photos by Dylan Toombs, Temerty Faculty of Medicine, Geoffrey Vendeville and Nick Iwanyshyn)

“You have four incredibly fierce and important Black women researchers moving us towards a thriving research culture,” says Coleman. “With their leadership, and with the university endorsing the Black Research Network as a strategic initiative, it makes it about more than addressing inequality and anti-Black racism – it’s also about shining a light on excellence through a directed, applied and resourced program.”

While the Black Lives Matter movement has recently put racism at the top of the daily news, the broad scope of the ISI initiative helps ensure that the Black Research Network can continue to make systemic change long after the news cycle has moved on.

The same goes for the work of the Emerging and Pandemic Infections Consortium, another ISI project.

“We don't want everyone to look away from infectious disease once the current pandemic recedes,” says Scott Gray-Owen, a professor of molecular genetics in the Temerty Faculty of Medicine who is director of the consortium. “There will be other pandemics, as well as more localized epidemics around the world. Toronto is a hub for transport and travel and is home to such an international community of people.

“Not only does that mean we’re likely to be affected, but we’re also a natural place to prepare to respond better next time.”

Scott Gray-Owen, a professor of molecular genetics in the Temerty Faculty of Medicine, is the director of the Emerging and Pandemic Infections Consortium (photo by Nick Iwanyshyn)

The consortium fosters innovative approaches to combatting infections through microbial biology, immunity and disease progression alongside the politics, economics and social impact of pandemic preparedness, prevention and mitigation.

“You’re trying to move society to a place where it actually prepares for something in advance,” Gray-Owen says. “And that is a very different type of challenge from just understanding microbial mechanisms.”

U of T has already created almost two dozen ISI projects, with more on the way. Each has its own structure and composition, but all take an evolved, advanced approach to mixing disciplines.

“Interdisciplinarity used to mean people would look at the same problem from different directions. That’s important and powerful, but we’re trying to deepen that integration. We’re all coming together to see how we can contribute to solving a problem,” Gray-Owen says. “We have experts around the table who are internationally renowned for their contribution to cancer or autoimmune diseases or surface decontamination. “The egos just go away, and new ideas come from that conversation.”

ISI projects also look beyond senior faculty and established scholars to engage students, government funders and policy-makers, commercial enterprises, not-for-profits, Indigenous groups, activists, citizen scientists, the media and the general public. The ISI networks typically also include training, mentoring and public education, which increases the impact of the research.

The programs aim to not only attract talent, but also grow it. Nejat, for one, says the openness of the ISI approach not only makes the research better, but it also helps ensure that the ideas that emerge have a positive impact in broader society.

“We've been very lucky to have a lot of interest in robotics from people who have never, ever dealt with the technology – people who come up with new research programs [related to robotics] that focus on the potential impact on society from a sociological or psychological point of view.

“It's really interesting to see them helping us pave that way into the future.”

This is the first article in a series about U of T’s Institutional Strategic Initiatives program – which seeks to make life-changing advancements in everything from infectious diseases to social justice – and the research community that’s driving it.

Climate change slows reduction of methylmercury levels in Arctic: U of T researchers

Christopher.Sorensen — Fri, 19 Feb 2021 18:31:22 +0000

Climate change slows reduction of methylmercury levels in Arctic: U of T researchers

Christopher.Sorensen Fri, 02/19/2021 - 13:31

Cutline

Igor Lehnherr, a researcher at U of T Mississauga, assessed the build-up of methylmercury, a dangerous neurotoxin, in Lake Hazen, one of Canada’s northernmost lakes (photo by Igor Lehnherr)

Topic

Research & Innovation

Sustainability

U of T Mississauga

Climate change may be slowing the reduction of methylmercury – a dangerous organic neurotoxin created by microbes that metabolize mercury – in Arctic waters despite a global movement to reduce industrial mercury emissions.

That is among the findings of Igor Lehnherr and his research team at the �ؿ�ζSM after assessing the build-up of methylmercury in Lake Hazen, one of Canada’s northernmost lakes.

The study is published in the journal Environmental Science & Technology.

“Mercury pollution has gone down in the atmosphere,” says Lehnherr, an assistant professor of geography at U of T Mississauga. “We’re doing things to tackle it, but climate change is throwing things for a loop [because] it can actually undo some of the benefits from emission reductions.”

Methylmercury levels rise only indirectly from human activity. Burning fossil fuels, mining and other industrial processes release unmethylated mercury into the atmosphere. As the mercury settles into aquatic ecosystems, certain types of microbes metabolize it to form the much more dangerous methylmercury.

A “persistent organic pollutant,” Methylmercury becomes more concentrated as it moves up the food chain – from bacteria to fish, predators and people. It affects the nervous system and can also cause cardiovascular damage. The toxin is especially dangerous for pregnant women and for fetuses, babies and young children whose nervous systems are still developing.

While the area where the U of T team collected samples is not close to any northern communities, Lehnherr says the work is relevant for Indigenous people who hunt and fish for food.

“What we’re learning is not constrained to that location,” he says. “We put a lot of import on understanding mechanisms that affect methylmercury, so we can apply what we learn in one place somewhere else.”

Igor Lehnherr, an assistant professor of geography at U of T Mississauga, says the field work for his latest study spanned several seasons and involved collaboration with other research teams in order to expand sampling (photo by Igor Lehnherr)

Arctic methylmercury levels depend on a complex mix of factors, including industrial emissions, precipitation patterns, microbial numbers and activity, as well as changes in seasonal sea ice. The complexity, along with the remoteness of northern ecosystems, make Lehnherr’s work particularly challenging.

“The field work spanned a few seasons,” he says of his latest study. “Some years we were there in the spring when it’s all snow and ice cover, some years in the summer, some years for both. By combining efforts with other teams, we expanded the sampling. Arctic research by nature is fairly collaborative – we share costs, time and ideas.”

Igor Lehnherr and his research team take water samples through the ice (photo by Igor Lehnherr)

In general, methylmercury-producing microbes are more active in warmer environments, implying a direct correlation between global warming and increased toxicity. But climate change also has many other effects that can exacerbate, mitigate and further complicate the situation.

“Temperature in the Arctic also controls permafrost thaw. It affects the amount of precipitation by controlling cloud cover, sea ice cover, rates of evaporation and these kinds of things,” Lehnherr says.

Changing weather patterns also affect how much methylmercury builds up in specific isolated areas and how efficiently it flows from one lake to the next, creating more widespread problems. In the short term, Lehnherr says it looks as though reduced emissions have not fully translated into cleaner Arctic ecosystems. However, Lehnherr says it should not necessarily be interpreted as a sign that efforts to reduce mercury aren’t worth it.

“I mostly think it validates the ongoing efforts to reduce anthropogenic mercury emissions,” he says. “Countries have shown this is something they’re willing to take on. These results allow us to have reasonable expectations about how long it will take for mercury levels to go down and stabilize.”

Lehnherr also wants to reassure people in northern communities who may be concerned about the safety of their food supply.

“Whenever I talk about the risks of mercury and negative health impacts, I always stress that the benefits of consuming traditional foods vastly outweigh the risks of contaminants. Locally caught Arctic char has better nutritional value than dried goods and flown-in goods,” he says.

Lehnherr plans to continue his study of methylmercury in the Arctic region to get a better sense of the long-term impacts of climate change.

Ice arches holding Arctic's ‘Last Ice Area’ in place are at risk, U of T researcher says

Christopher.Sorensen — Tue, 05 Jan 2021 14:21:49 +0000

Ice arches holding Arctic's ‘Last Ice Area’ in place are at risk, U of T researcher says

Christopher.Sorensen Tue, 01/05/2021 - 09:21

Cutline

Sea ice in the Nares Strait as seen from a NASA P-3B turboprop during a 2013 survey flight (photo by Christy Hansen/NASA)

Topic

Research & Innovation

U of T Mississauga

Snugged up against the upper edges of the Canadian Arctic Archipelago and Greenland lies the oldest and thickest sea ice in the world, covering hundreds of thousands of square kilometres of ocean. Arctic sea ice grows and shrinks with the seasons, but this ice has so far lasted even through the warmest summers on record.

Scientists call this region “The Last Ice Area.” They say it could endure even after the rest of the Arctic becomes ice-free in the warmer months, providing a vital refuge for polar bears, walruses and other species that rely on sea ice to survive.

But recent research at the �ؿ�ζSM Mississauga suggests the Last Ice Area may be in more peril than previously thought. In a recent paper published in the journal Nature Communications, Professor Kent Moore and his co-authors describe how this multi-year ice is at risk not just of melting in place, but of floating southward into warmer regions. This, in turn, would create an “ice deficit” and hasten the disappearance of the Last Ice Area.

“This very old ice is what we’re concerned about,” says Moore, who is in U of T Mississauga’s department of chemical and physical sciences. “The hope is that this area will persist into the middle part of this century or even longer. And then, hopefully, we'll eventually be able to cool the planet down. The ice will start growing again, and then this area can act as a sort of seed,”

Using satellite data, Moore has been studying ice arches that form along Nares Strait, a 40-kilometre-wide, 600-kilometre-long channel that runs between Greenland and Ellesmere Island from the Arctic Ocean into Baffin Bay.

Moore had already observed warning trends in earlier research that indicated this ice is increasingly on the move.

“The Last Ice Area is losing ice mass at twice the rate of the entire Arctic,” Moore says. “We realized this area may not be as stable as people think.”

His most recent analysis of satellite data says the problem may be getting even worse. The arches along Nares Strait that historically have held the Last ice Area in place have become less stable, according the study.

“The ice arches that usually develop at the northern and southern ends of Nares Strait play an important role in modulating the export of Arctic Ocean multi-year sea ice,” he and his authors write.

“The duration of arch formation has decreased over the past 20 years, while the mass of ice exported through Nares Strait has increased.”

The ice arches form as the weather cools. Multiple ice floes converge as they funnel into the relatively narrow strait, forming huge structures that look like bridge supports turned on their sides. The arches span the full width of the passage, blocking the movement of multi-year ice from north to south.

“It's really quite profound to imagine a 100-kilometre-long barrier of ice that remains stationary for months at a time. That's more than twice as long as Louisiana’s Lake Pontchartrain Causeway – the world’s longest continuous bridge over water,” Moore says. “It speaks to the strength of ice.”

But that strength is diminishing. Ice arches only form for part of the year. When they break up in the spring, ice moves more freely down the Nares Strait. And that breakup is happening sooner than in the past.

“Every year, the reduction in duration is about one week,” Moore says. “They used to persist for about 200 days and now they’re persisting for about 150 days. There’s quite a remarkable reduction.

“We think that it’s related to the fact the ice is just thinner and thinner ice is less stable.”

The impact of losing the Last Ice Area would extend far beyond photogenic species like polar bears. Ice algae flourishes below the ice and in brine channels that run through its cracks and fissures, supplying carbon, oxygen and nutrients that underpin an elaborate but vulnerable ecosystem.

In 2019, the Canadian government designated a section of the Last Ice Area as the Tuvaijuittuq Marine Protected Area. Tuvaijuittuq is Inuktut for “the place where the ice never melts.”

Moore remains hopeful that his analysis of the Nares Strait ice arches will focus more attention on this important region of the Arctic. However, he says action targeted specifically at preserving the arches won’t be sufficient to solve the problem. A global solution is needed.

“The scale is so huge and the region is so remote,” he says. “The only thing we can do is cool the planet down. Then the arches will hopefully naturally form again.”

Inset: Images of the ice arch that formed at the southern end of Nares Strait in 2020. The upper image shows the arch holding back the ice while the lower image shows the ice streaming southwards after the arch collapses (Sentinel-2 satellite imagery courtesy of the European Space Agency)

Gut feeling: U of T Mississauga research reveals how honeybees identify outsiders

Christopher.Sorensen — Tue, 27 Oct 2020 14:47:36 +0000

Gut feeling: U of T Mississauga research reveals how honeybees identify outsiders

Christopher.Sorensen Tue, 10/27/2020 - 10:47

Cutline

(photo by Damien Tupinier via Unsplash)

Patchen Barss

Topic

Breaking Research

Biology

Research & Innovation

U of T Mississauga

Honeybees identify strangers in their midst using signals from the bacteria in the intruder’s digestive system, an international team of biologists has found.

The research, published recently in Science Advances, contributes to broader questions about the tools social animals use to distinguish between insiders and outsiders in their community.

“This idea of how an individual knows another individual, how they recognize whether they're like themselves or not, is a mystery,” says Professor Joel Levine, chair of biology at U of T Mississauga and one of the paper’s authors.

Many animals distinguish members and non-members of their pack, herd or flock using visual cues, scents, behaviours, genetic relationships and other factors.

Bee colonies, with their sophisticated, hierarchical structures, “are the best show in town when it comes to sociality,” Levine says.

The social insects take on different roles, including tending the queen, caring for young, cleaning the hive, building honeycomb, collecting nectar and pollen and producing honey. Bees have no trouble recognizing the other members of their hive. But should invaders from another nest try to infiltrate, guard bees quickly identify the stranger and attack.

The research team was interested in how that identification happens. Earlier research had eliminated differences in genetic code as the main mechanism.

“It's a very common beekeeping technique to take newly emerged bees from a strong colony and dump them into a weak colony to make that weaker colony stronger,” says Cassondra Vernier, a biologist at Washington University who developed the microbiome theory at the heart of their research. “Those bees will actually grow up in this new colony that they're not genetically related to. And they will develop the cue that matches that colony. The bees of that colony will recognise them as nestmates.”

Vernier theorized that bee-belly bacteria might hold answers. She hypothesized that, in the close proximity of hive life, bees would exchange bacteria, creating a common “microbiome” across the colony.

She shipped frozen bee remains from her laboratory in St. Louis, Mo. to Mississauga, Ont., where Levine and Joshua Krupp, a research associate, analyzed variations in the insects’ microbiomes. They confirmed that bees from each colony shared common gut bacteria, distinct from those from other hives. In addition, the distinct microbiome changed the type of pheromones the bees produced. Hive members exude a unique combination of chemicals, making a kind of signature that other bees can sense — likely through a combination of taste, smell and texture.

The researchers are cautious about assigning human-like abilities to bees’ perceptual capacities. In some ways, the us-and-them behaviour in the bee world has more in common with the way immune cells identify and attack invaders than the way thinking organisms treat outsiders. But there’s still much to learn about the boundaries between chemical and social reactions.

“We’re a long way from making an argument that an individual bee is a self-conscious, sentient being,” says Levine. “But we’re not so far from saying that these guys know something about each other, and that they're making decisions based on that knowledge when they interact. Does that have evolutionary implications that may affect how we understand other species all the way up to and including humans? I would say absolutely. But can we say yet how it's going to shake out? No way.”

The research received support from the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council, among others.

Can we eliminate bias in AI? How Canada’s commitment to multiculturalism could help it become a world leader

noreen.rasbach — Wed, 24 Jul 2019 15:03:25 +0000

Can we eliminate bias in AI? How Canada’s commitment to multiculturalism could help it become a world leader

noreen.rasbach Wed, 07/24/2019 - 11:03

Cutline

(illustration by Sébastien Thibault)

Patchen Barss

Topic

Global Lens

Artificial Intelligence

Computer Science

Diversity

Faculty of Arts & Science

Faculty of Medicine

Global

Research & Innovation

Vector Institute

Subheadline

In human beings, intelligence is no inoculation against bias and bigotry. The same holds true for computers. Intelligent machines learn about the world through the filters of human language and historical behaviour – meaning they can just as easily absorb humanity’s worst values as they can its best.

Researchers who aim to develop ever-smarter machines have their work cut out for them to ensure that they’re not inadvertently imbuing computers with misogyny, racism or other forms of bigotry.

“It’s a huge risk,” says Marzyeh Ghassemi, an assistant professor in the �ؿ�ζSM's department of computer science and Faculty of Medicine who focuses on health-care applications for artificial intelligence (AI). “Like all advances that leapfrog societies forward, there are large risks that we must decide to accept or not to accept.”

Bias can creep into algorithms in many ways. In a highly influential branch of AI known as “natural language processing,” problems can arise from the “text corpus” – the source material the algorithm uses to learn about the relationships between different words.

Natural language processing, or “NLP,” allows a computer to understand human-style speech – informal, conversational and contextual. NLP algorithms comb through billions of words of training text – the corpus might be, say, the entirety of Wikipedia. One algorithm works by assigning to each word a set of numbers that reflects different aspects of its meaning – “king” and “queen” for instance, would have similar scores relating to the idea of royalty, but opposite scores relating to gender. NLP is a powerful system that allows machines to learn about relationships between words – in some cases, without direct human involvement.

“Even though we’re not always teaching them specifically, what they learn is incredible,” says Kawin Ethayarajh, a researcher who focuses partly on fairness and justice in AI applications. “But it’s also a problem. In the corpus, the relationship between ‘king’ and ‘queen’ might be similar to the relationship between ‘doctor’ and ‘nurse.’”

But of course, all kings are men; not all doctors are men. And not all nurses are women.

When an algorithm absorbs the sexist tropes of historical human attitudes, it can lead to real-life consequences, as happened in 2014 when Amazon developed an algorithm to vet job applicants’ resumés. The company trained its machines using 10 years of hiring decisions. But in 2015, they acknowledged that, in tests, the system was giving unearned preference to resumés from male applicants. They tweaked the system to force it to ignore gender information, but ultimately shut down the project before actually putting it to use as they could not be sure their algorithm wasn’t perpetrating other forms of discrimination.

Mitigating sexist source material can involve technological and methodological adjustments. “If we can understand exactly what underlying assumptions the corpus has that cause these biases to be learned, we can either select corpora without those biases or correct it during the training process,” says Ethayarajh.

It’s common practice for researchers to design an algorithm that corrects prejudicial assumptions automatically. By adjusting the weight of the numbers it assigns to each word, the computer can avoid making sexist or racist associations.

But what exactly are the assumptions that need correcting? What does a fair-minded AI really look like? Debates over privilege, bigotry, diversity and systemic bias are far from settled. Should a hiring algorithm have a stance on affirmative action? Should a self-driving car take special care if another vehicle has a “Baby on Board” sticker? How should an AI-driven analysis of legal documents factor in the historical treatment of Indigenous Peoples? Contentious societal issues don’t disappear merely because machines take over certain recommendations or decisions.

Many people view Canada’s flawed but relatively successful model of multiculturalism as a chance to lead in fair AI research.

“Canada certainly does have an opportunity,” says Ronald Baecker, a professor emeritus of computer science and the author of Computers and Society: Modern Perspectives. He sees a role for government to redress the societal inequities, injustices and biases associated with AI by, for example, setting up protections for employees who choose to speak out against biased or unfair AI-driven products. “There’s a need for more thinking and legislation with respect to the concept of what I would call ‘conscientious objection’ by high-tech employees.”

He also believes that the computer scientists developing smart technologies should be required to study the societal impact of such work. “It’s important that professionals who work in AI recognize their responsibility,” he says. “We’re dealing with life-and-death situations in increasingly important activities where AI is being used.”

Algorithms that help judges set bail and sentence criminals can absorb long-standing biases in the legal system, such as treating racialized people as if they are more likely to commit additional crimes. The algorithms might flag people from certain communities as posing too high a risk to receive a bank loan. They also might be better at diagnosing skin cancer in white people than in people with darker skin, as a result of having been trained on skewed source material.

The stakes are incredibly high in health care, where inequitable algorithms could push people who have been poorly served in the past even further into the margins.

In her work at U of T and at the Vector Institute for Artificial Intelligence, Ghassemi, like other researchers, takes pains to identify potential bias and inequity in her algorithms. She compares the recommendations and predictions of her diagnostic tools against real-world outcomes, measuring their accuracy for different genders, races, ages and socio-economic factors.

“Like all advances that leapfrog societies forward, there are large risks that we must decide to accept or not to accept,” says Marzyeh Ghassemi, an assistant professor in the department of computer science who focuses on health-care applications for artificial intelligence

In theory, Canada offers a head start for researchers interested in health-care applications that reflect values of fairness, diversity and inclusion. Our universal health-care system creates a repository of electronic health records that provides a wealth of medical data that could be used to train AI-driven applications. This potential drew Ghassemi to Toronto. But the technology, information, formatting and rules to access these records vary from province to province, making it complicated to create the kind of data sets that can move research forward.

Ghassemi was also surprised to learn that these records only rarely include data about race. This means if she’s using an algorithm to determine how well a given treatment serves different sectors of society, she could identify disparities between men and women, for example, but not between white people and racialized people. As a result, in her teaching and research, she’s using publicly available American data that contains information about race.

“Auditing my own models [using American data], I can show when something has higher inaccuracy for people with different ethnicities,” she says. “I can’t make this assessment in Canada. There’s no way for me to check.”

Ghassemi is interested in creating AI applications that are fair in their own right – and that also can help human beings counteract their own biases. “If we can provide tools based on large diverse populations, we’re giving doctors something that will help them make better choices,” she says.

Women, for example, are significantly underdiagnosed for heart conditions. An AI could flag such a danger for a doctor who might overlook it. “That’s a place where a technological solution can help, because doctors are humans, and humans are biased,” she says.

Ethayarajh concurs with Ghassemi and Baecker that Canada has an important opportunity to press its advantage on fairness and bias in artificial intelligence research.

“I think AI researchers here are very aware of the problem,” Ethayarajh says. “I think a part of that is, if you look around the office, you see a lot of different faces. The people working on these models will be end-users of these models. More broadly, I think there is a very strong cultural focus on fairness that makes this an important area for researchers in this country.”

This story originally appeared in �ؿ�ζSM Magazine.

U of T physicists discovered a way to increase the resolution of microscopes and telescopes

ullahnor — Thu, 16 Feb 2017 19:40:35 +0000

U of T physicists discovered a way to increase the resolution of microscopes and telescopes

ullahnor Thu, 02/16/2017 - 14:40

Cutline

U of T PhD students Edwin (Weng Kian) Tham and Hugo Ferretti are part of the team that helped develop a way to look at other properties of light (photo by Diana Tyszko)

Patchen Barss

Author legacy

Patchen Barss

Topic

Faculty of Arts & Science

�ؿ�ζSM researchers have found a way to increase the resolution of microscopes and telescopes beyond long-accepted limitations by tapping into previously neglected properties of light.

The method allows observers to distinguish very small or distant objects that are so close together they normally meld into a single blur.

The research appears in the journal Physical Review Letters.

Because of the laws of physics, which cause light to spread out or “diffract,” telescopes and microscopes are great for observing lone subjects. With an object like a binary star on the other hand, two stars that are close together may appear at a distance as one blurry dot, and their individual information is irrevocably lost.

Part of the problem is circumventing the limitations of what is referred to as the “Rayleigh Criterion.”

More than 100 years ago, British physicist John William Strutt – better known as Lord Rayleigh – established the minimum distance between objects necessary for a telescope to pick out each individually. The “Rayleigh Criterion” has stood as an inherent limitation of the field of optics ever since.

Telescopes, though, only register light’s “intensity” or brightness. Light has other properties that now appear to allow one to circumvent the Rayleigh Criterion.

“To beat Rayleigh’s curse, you have to do something clever,” says Professor Aephraim Steinberg, a physicist at U of T’s Centre for Quantum Information and Quantum Control and senior fellow in the quantum information science program at the Canadian Institute for Advanced Research.

“We measured another property of light called ‘phase.’ And phase gives you just as much information about sources that are very close together as it does those with large separations.”

Light travels in waves, and all waves have a phase. Phase refers to the location of a wave’s crests and troughs. Even when a pair of close-together light sources blurs into a single blob, information about their individual wave phases remains intact. You just have to know how to look for it.

This realization was published by National University of Singapore researchers Mankei Tsang, Ranjith Nair, and Xiao-Ming Lu last year in Physical Review X. Researchers like Steinberg and his team immediately set about devising a variety of ways to put it into practice.

“We tried to come up with the simplest thing you could possibly do,” Steinberg says. “To play with the phase, you have to slow a wave down, and light is actually easy to slow down.”

His team, including PhD students Edwin (Weng Kian) Tham and Hugo Ferretti, split test images in half. Light from each half passed through glass of a different thickness, which slowed the waves for different amounts of time, changing their respective phases. When the beams recombined, they created distinct interference patterns that told researchers whether the original image contained one object or two – at resolutions well beyond the Rayleigh Criterion.

So far, Steinberg’s team has tested the method only in artificial situations involving highly restrictive parameters.

“I want to be cautious – these are early stages,” Steinberg says. “In our laboratory experiments, we knew we just had one spot or two, and we could assume they had the same intensity. That’s not necessarily the case in the real world. But people are already taking these ideas and looking at what happens when you relax those assumptions.”

The advance has potential applications both in observing the cosmos, and also in microscopy, where the method can be used to study bonded molecules and other tiny, tight-packed structures.

Regardless of how much phase measurements ultimately improve imaging resolution, Steinberg says the experiment’s true value is in shaking up physicists’ concept of “where information actually is.”

Steinberg’s “day job” is in quantum physics – this experiment was a departure for him. He says work in the quantum realm provided key philosophical insights about information itself that helped him beat “Rayleigh’s curse.”

“When we measure quantum states, you have something called the Uncertainty Principle, which says you can look at position or velocity, but not both,” he says. “You have to choose what you measure. Now we’re learning that imaging is more like quantum mechanics than we realized. When you only measure intensity, you’ve made a choice, and you’ve thrown out information. What you learn depends on where you look.”

Support for the research was provided by by the Natural Sciences and Engineering Research Council of Canada, the Canadian Institute for Advanced Research, and Northrop-Grumman Aerospace Systems NG Next.

Team led by U of T researchers discovers energy source sustaining microbial life deep beneath Earth’s surface

ullahnor — Thu, 27 Oct 2016 15:49:09 +0000

Team led by U of T researchers discovers energy source sustaining microbial life deep beneath Earth’s surface

ullahnor Thu, 10/27/2016 - 11:49

Cutline

Sulfide minerals in the host rock, including pyrite, are oxidized by products of radiolysis to produce the source of sulfate found in the fracture waters (photos by K. Gorra)

Patchen Barss

Author legacy

Patchen Barss

Topic

Global Lens

Faculty of Arts & Science

Earth Sciences

Barbara Sherwood Lollar

Research

Collaboration

In Northern Ontario, a team led by U of T researchers has found the “geochemical fingerprints of life” and the energy sustaining this life in waters more than two kilometres below the surface of the Earth. The discovery demonstrates how life can be sustained even in the seemingly inhospitable environments of the deep Earth crust.

Most life on Earth gets its energy – directly or indirectly – from the sun. But there are other options.

“Microbial subsurface communities are often chemosynthetic, not photosynthetic,” says University Professor Barbara Sherwood Lollar in the department of earth sciences at the Faculty of Arts & Science. “In chemosynthesis, a molecule like hydrogen ‘donates’ electrons, and sulfate ‘accepts’ them. Basically, all metabolism works through this kind of exchange of electrons. That’s how energy works. That’s how life works.”

The chemical reactions producing the electron donor in these deep waters had been identified several years ago, but the source of sulfate – the electron acceptor – had been elusive.

Co-authors Barbara Sherwood Lollar and Georges Lacrampe-Couloume, holding pyrite rich rocks from the field site ( photo by K. Gorra)

In a paper published this week in Nature Communications, Sherwood Lollar and her colleagues report that sulfate dissolved in these waters 2.4 km below the surface comes from oxidation of the sulfide minerals in the ancient rocks via chemicals produced when radiation breaks the water down into its constituent parts.

First author Long Li, now the Canada Research Chair in stable isotope geochemistry at the University of Alberta, worked with Sherwood Lollar at U of T as a postdoctoral fellow. Along with researchers from McGill University, they studied the distribution pattern of multiple sulfur isotopes – that is, sulfur atoms that differ by the number of neutrons – in the dissolved sulfate in ancient subterranean waters near Timmins, Ont.

Read about the latest findings in Deep Carbon Observatory

Their earlier work had revealed that these waters contain hydrogen and sulfate – key components that make life possible without sunlight. The multiple sulfur isotope compositions in the sulfate show a unique pattern, only seen in rocks formed before oxygen appeared in Earth’s atmosphere about 2.4 billion years ago.

By matching this isotopic feature in the dissolved sulfate with that of pyrite in the 2.7-billion-year-old rocks hosting the waters, the researchers demonstrated that the same pyrite and other sulfide ores that make these rocks ideal for economic mining of metals, produce the “fuel” for microbial metabolisms.

But there were other surprises in store.

“When we looked at the sulfate dissolved in these waters, we found it was more enriched in an isotope called sulphur 34 than expected,” Sherwood Lollar says.

Read about Barbara Sherwood Lollar being honoured by the Royal Society of Canada

Making waves: How the �ؿ�ζSM made the discovery of gravitational waves possible

lavende4 — Tue, 04 Oct 2016 18:45:28 +0000

Making waves: How the �ؿ�ζSM made the discovery of gravitational waves possible

lavende4 Tue, 10/04/2016 - 14:45

Cutline

Left to right: PhD candidate Heather Fong, CITA professor Harald Pfeiffer and CITA post doctoral fellow Prayush Kumar (Photo by Diana Tyszko)

Patchen Barss

Author legacy

Patchen Barss

Topic

Global Lens

LIGO

Gravitational waves

Faculty of Arts & Science

Canadian Institute for Theoretical Astrophysics

The discovery of gravitational waves in 2016 – predicted by Albert Einstein – was one of the biggest sciences stories of the decade. While the discovery didn’t take the Nobel Prize today, many in the scientific community think it is just a matter of time before the Nobel committee honours the Laser Interferometer Gravitational-Wave Observatory (LIGO) team for one of the most outstanding contributions in the field of physics. Science writer Patchen Barss explains the discovery and the key contributions of �ؿ�ζSM’s Harald Pfeiffer and his team.

A billion years ago and a billion light years away, a star 36 times more massive than our Sun expended its remaining fuel in a final blast of nuclear fusion. With nothing left to burn, the star began to collapse under its own gravity. The atoms in its massive core collapsed like crushed soda cans. Protons and electrons ground together to form new neutrons.

The star’s density kept increasing. Its gravity became so concentrated and intense that not even light could escape any longer. Spacetime warped and ruptured. The star became a black hole.

But that wasn’t the end of the story.

A second black hole, the product of an only slightly smaller stellar cataclysm passed by. The two became trapped in each other’s mighty gravitational fields. They circled one another, slowly at first, but then more and more quickly. Their collision course became a high-speed death spiral that sent waves of gravitational energy rippling out across their galaxy and into the cosmos at the speed of light, stretching and squeezing space itself.

Back on present-day Earth, came a different kind of merger: a collision of ideas between observational cosmologists and numerical relativity experts.

In the 1970s, observational scientists had begun working on “laser interferometry” instruments that might detect gravitational waves. Decades of effort culminated in the construction of the Laser Interferometer Gravitational-Wave Observatory (LIGO), which comprises two massive detectors, one in Washington and the other in Louisiana.

Concurrently, the �ؿ�ζSM was leading an international effort to simulate black hole collisions and predict what the emerging gravitational wave patterns might look like.

The simulations belong to a field of study with the unglamorous name of “numerical relativity.” These supercomputer simulations nab few headlines, but without them, gravitational-wave research wouldn’t have gone far, even with LIGO’s whiz-bang technology. Scientists at U of T identified the need for powerful simulations early in LIGO’s planning stages, and drove the push to mature the theoretical science in time to make the most of LIGO’s observations.

“The development of these simulations was precisely designed to make us able to analyze the data collected by LIGO experiments,” says J. Richard Bond, a University Professor at the Canadian Institute for Theoretical Astrophysics (CITA) in the Faculty of Arts & Science. Bond drove the effort to recruit an expert devoted to numerical relativity.

“Detecting gravitational waves is a huge revolution. It will be front and centre in what’s going to happen over the next few decades,” he says. “You’re either on that bus or off it. Somebody here at the University had to be on the gravitational-wave bus.”

In fact, U of T attracted a whole busload of graduate students and postdoctoral fellows to work on numerical relativity. From the start, though, the person driving that bus has been Harald Pfeiffer.

Before Pfeiffer became an associate professor at CITA, he had already established his reputation in numerical relativity at Cornell University and Caltech.

“I have always been interested in black holes and Einstein and gravity and computers,” he says. “At Cornell, I worked with one of the world’s experts on solving Einstein’s equations on supercomputers. The relevance to LIGO was there all along.”

In the early 20^th century, Albert Einstein proposed his Theory of Relativity a model of gravity and the universe that scientists have been testing and exploring ever since. Many non-scientists can recite Einstein’s most famous equation: E=mc². But the so-called mass-energy equivalence equation is just one tiny part of the math behind relativity. Researchers are still finding new predictions based on Einstein’s equations, and using them to understand and simulate cosmic events that would otherwise defy imagination and intuition.

“The first time people tried to simulate black-hole collisions on computers was in 1964,” says Pfeiffer. “But even when I started my PhD, nobody had yet figured out how to do it. We made steady progress but only on arcane technical sub-problems. The big problem eluded everybody until 2005 when finally all the pieces came together.”

LIGO faced hurdles of its own. Through the 1980s and 1990s, the project faced technological and budgetary delays. Between 2002 and 2010, the first major version of LIGO worked exactly as planned. But, during that time the cosmos failed to cooperate, sending no detectable waves our way. An international team of scientists continued to make refinements and improvements to increase LIGO’s sensitivity.

LIGO’s L-shaped detectors work by splitting a laser beam into two waves radiating at right angles to one another. Each beam travels precisely the same distance – four kilometres – through a vacuum, bounces off a fine-tuned mirror and returns along the same path to the split point. In the absence of gravitational waves, the returning beams cancel each other out. The detector stays quiet.

But passing gravitational waves would lengthen space in one direction and squeeze it in the other. Each beam would travel a slightly different distance, get out of sync with the other, and create a distinct, detectable interference pattern.

Researchers built two such detectors thousands of kilometres apart, which allowed them not just to detect waves, but also to triangulate them to determine the location of their source.

They still needed to know what to look for, though.

“In the first 10 years, my research and LIGO were not directly touching each other,” says Pfeiffer. “However, on both sides of the fence there was momentum building and building rapidly.”

Both sides were working toward a goal that nobody was sure would be achievable. Still, they were spiraling in on one another, circling toward an explosive discovery.

In September 2015 a new, vastly more sensitive iteration of LIGO came online. By then, Pfeiffer and his team had simulated thousands of collisions, creating a bank of “pattern templates” that gave observers clues about what to look for, and how to interpret what they found.

Not long after, gravitational waves from that distant, ancient black-hole collision finally reached the Earth.

Space compressed in one direction, stretched in another. The laser beams fell out of sync.

“Chirp!”

That chirp, revealed to the world at an international press conference in February 2016, was an audio interpretation of a laser interference pattern created by billion-year-old gravitational waves.

Using Pfeiffer’s simulations, researchers conclusively identified the pattern as the first-ever direct detection of gravitational waves.

“U of T’s key contribution was this waveform modeling,” says Pfeiffer. “If you know the shape of the signal you’re looking for, it’s like knowing the colour of a needle in a haystack. It’s easier to find.”

In June 2016, LIGO scientists announced that the detectors had chirped again: a second detection. In this case, though, the black holes involved had about one third the combined mass of the first collision. It was a “quieter” crash with a weaker signal, which meant the simulations played an even more important role in its interpretation.

“The second detection would have been an extremely marginal discovery without the simulations,” says Pfeiffer. “It would have been flagged as an interesting detection, possibly between two black holes, but nothing more precise.”

The pattern templates also save time – rather than deciphering data for days on end, observers can say right away, “You’ve got waves!”

“Real-time is important, because there’s a whole band of astronomers across the world who are not part of LIGO,” says Peter Martin, a CITA professor. “They want to turn optical or radio telescopes to the point of detection quickly to see whether any electromagnetic flash comes with gravitational radiation.”

Researchers continue to improve LIGO, with plans to double its sensitivity. That puts pressure on Pfeiffer to keep building simulations based on Einstein’s relativity equations.

“As boring as it sounds, there’s still a lot of work to be done in improving the waveforms that LIGO is looking for,” says Pfeiffer. “It’s really cool having this big breakthrough, but 99 percent of science is the tedious day-to-day work.”

LIGO plans to continue observations in 2016, and it will join forces with a French-Italian gravitational-wave detector in 2017. Plans include studying more colliding black holes, scoping out their properties in unprecedented detail, and checking whether Einstein’s theory continues to work flawlessly in light of ever more precise data.

Astronomers will also search for gravitational waves from sources other than black holes, including from less-massive-but-still-whoppingly-massive bodies like pulsars and other neutron stars that spin at high speed.

Bond, though, has his eye on another target.

“In Toronto, I and many others are heavily invested in discovering gravitational waves formed during the first moments of the universe,” Bond says.

“The sheer challenge of figuring out how to solve Einstein’s equations would have been enticing enough of a problem,” Pfeiffer says. But he found it doubly exciting when those equations allowed scientists to precisely reconstruct the story of that distant, cataclysmic collision from a billion years earlier.

“It is amazingly satisfying, to see the effort of thousands of people come together,” he says. “Building the LIGO instruments, developing the software to analyse the data, and also our own contribution toward detecting and deciphering the signals. It was only through this huge joint effort that we could discover black holes colliding.”

Didn't see that stop sign? U of T researchers examine the invisible world of human perception

sgupta — Mon, 21 Mar 2016 06:50:07 +0000

Didn't see that stop sign? U of T researchers examine the invisible world of human perception sgupta Mon, 03/21/2016 - 02:50

Cutline

graduate student Jason Rajsic, postdoctoral fellow Eric Taylor and Professor Jay Pratt of psychology (photo by Diana Tyszko)

Patchen Barss

Author legacy

Patchen Barss

Topic

Breaking Research

Features

Faculty of Arts & Science

Research

Stage magicians are not the only ones who can distract the eye: a new cognitive psychology experiment from the �ؿ�ζSM demonstrates how all human beings have a built-in ability to stop paying attention to objects that are right in front of them.

We see much less of the world than we think we do

Perception experts have long known that we see much less of the world than we think we do. A person creates a mental model of their surroundings by stitching together scraps of visual information gleaned while shifting attention from place to place. Counterintuitively, the very process that creates the illusion of a complete picture relies on filtering out most of what’s out there.

In a paper published in the journal Attention, Perception, & Psychophysics a team of U of T researchers reveal how people have more “top-down” control of what they don’t notice than many scientists previously believed.

“The visual system really cares about objects,” says postdoctoral fellow J. Eric T. Taylor, who is the lead author on the paper. “If I move around a room, the locations of all the objects — chairs, tables, doors, walls, etc. — change on my retina, but my mental representation of the room stays the same.”

Objects play fundamental role in how we focus our attention

Objects play such a fundamental role in how we focus our attention that many perception researchers believe we are “addicted” to them; we couldn’t stop paying attention to objects if we tried. The visual brain guides attention largely by selecting objects — and this process is widely believed to be automatic.

“I had an inkling that object-based attention cues require a little more will on the observer’s part,” says Taylor. “I designed an experiment to determine whether you can ‘erase’ object-based attention shifting.”

Taylor put a new twist on an old and highly influential test known as a “two-rectangle experiment.” The original experiment was instrumental in demonstrating just how deeply objects are ingrained in how we see the world.

In the original experiment, test subjects stare at a screen with two skinny rectangles. A brief flash of light draws their attention to one end of one rectangle — say the top end of the left rectangle. Then, a “target” appears, either in the same place as the flash, at the other end of the same rectangle, or at one of the ends of the other rectangle.

Observers are consistently faster at seeing the target if it appeared at the opposite end of the original rectangle than if it appeared at the top of the other rectangle — even though those two points are precisely the same distance from the original flash of light.

The widely accepted conclusion was that the human brain is wired to use objects like these rectangles to focus attention. Alternately referred to as a “bottom-up” control or a “part of our lizard brain,” object-based attention cues seemed to evoke an involuntary, uncontrolled response in the human brain.

New element added to research: colour

Taylor and colleague’s variations added a new element: test observers went through similar exercises, but they were instructed to hunt targets of a specific colour that either matched or contrasted with the colour of the rectangles themselves.

“They activate a ‘control setting’ for, say, green, which is a very top-down mental activity,” says Taylor. “We found that when the objects matched the target colour, people use them to help direct their attention. But when the objects were not the target colour, people no longer use them — they become invisible.”

Test observers are aware of the rectangles on the screen, but when they’re seeking a green target among red shapes, those objects no longer affect the speed with which they find it. In everyday life, we continually create such top-down filters, by doing anything from heeding a “Watch for children” sign to scanning a crowd for a familiar face.

“This result tells us that one of the ways we move attention around is actually highly directed rather than automatic,” Taylor says. “We can’t say exactly what we’re missing, but whatever is and is not getting through the filter is not as automatic as we thought.”

picpath

sites/default/files/2016-03-21-perception.jpg

U of T physicists taking a cold look at atomic interactions

sgupta — Mon, 22 Feb 2016 15:23:57 +0000

U of T physicists taking a cold look at atomic interactions sgupta Mon, 02/22/2016 - 10:23

Cutline

Diana Tyszko.)

Patchen Barss

Author legacy

Patchen Barss

Topic

Subheadline

Measurements of “p-waves” will aid understanding of superconductivity and other fundamental material properties

Researchers led by Joseph Thywissen of the �ؿ�ζSM’s department of physics have discovered new rules governing the behaviour of atoms that collide at temperatures close to absolute zero.

By adjusting the magnetic field surrounding these atoms and making observations through a process called dynamical spectroscopy, the researchers were able to measure correlations and find evidence for the laws governing p-wave interactions, which are rare in nature. The result could be new insights into superconductivity, superfluidity and other fundamental properties of materials.

The research appears in a paper published on Feb. 22 in the journal Nature Physics.

“Ultracold atoms are the stem cells of materials science,” says Thywissen, a professor of physics and fellow of the quantum materials program at the Canadian Institute for Advanced Research.

“Just as a stem cell can become a fingernail or a heart cell depending on its context, ultracold atoms can become metals, insulators, superfluids or other types of materials.”

In collaboration with theorists Shizhong Zhang of Hong Kong University and Zhenhua Yu of Tsinghua University, the Toronto scholars have been studying p-wave interactions in a highly controlled environment, coaxing a few hundred thousand gas atoms into a “trap,” and cooling them close to absolute zero (-273.15 Celsius).

If two atoms hit head-on and bounce straight back from one another, they have no angular momentum. This interaction is called an s-wave. But if a pair of atoms ricochet off one another with a single unit of angular momentum, the resulting interaction is known as a p-wave.

P-waves, s-waves and other types of atom-pair interactions correlate with many types of emergent physical properties. Some rules that govern these relationships are well understood, but those related to p-waves have traditionally defied explanation.

“P-wave interactions fascinate scientists because they endow materials with unusual properties and puzzling behaviours,” says Thywissen. “But the conventional wisdom was that gases with p-wave interactions had losses that were too strong to allow you see anything interesting.”

Thywissen’s team employed dynamical spectroscopy to prepare and probe atoms faster than had been done in the past.

“Our observations took less than a millisecond,” he says. “Previous studies were searching for properties that required longer observation. It allowed us to see something before the losses became too significant.”

Their orthodoxy-challenging experiments resulted more from serendipity than a conviction that there was a problem with conventional wisdom.

“We ended up looking at this because a graduate student working in our lab didn’t know how to avoid the p-wave resonances,” Thywissen says. “He took spectroscopy data on them. Nature surprised us. There was a beautiful spectroscopic signal of a new kind of pressure that was due to p-wave interactions.”

Their subsequent observations sparked a flurry of activity among theoretical physicists, resulting in several papers that attempted to explain this pressure. If correct, this theoretical work provides a new set of guidelines outlining how to understand any state of matter that emerges from p-wave interactions.

This work can help scientists better understand the fundamental question of where material properties come from. It can also make it possible to create and work with new materials that have highly unusual – and potentially very valuable – properties.

P-waves, for instance, correlate with unusual forms of superconductivity and superfluidity, in which particles flow without resistance. Such materials have vexed scientists for years.

“When made up of p-wave pairs, superconductors and superfluids should also have something called an edge current – but we know from observation that these edge currents are absent or extremely weak,” says Thywissen. “We don’t understand this. We hope the new relations we’ve discovered will help us figure out why.”

Thywissen and his collaborators are already designing new experiments designed to tune and tweak the environment, creating an even more sophisticated understanding of how material properties emerge.

“Even though this experiment looks complex now, we will continue to work to push the limits of what can be done in the lab,” Thywissen says. “We never know what we’re going to find, but we always have hope of discovering something like this. It is truly thrilling.”

Research was carried out at U of T by PhD candidates Christopher Luciuk and Scott Smale, and postdoctoral fellow Stefan Trotzky, in addition to Yu, Zhang and Thywissen.

picpath

sites/default/files/p-waves.jpg

Patchen Barss

Groundbreakers: U of T initiative brings together experts to address major societal issues

Climate change slows reduction of methylmercury levels in Arctic: U of T researchers

Ice arches holding Arctic's ‘Last Ice Area’ in place are at risk, U of T researcher says

Gut feeling: U of T Mississauga research reveals how honeybees identify outsiders

Can we eliminate bias in AI? How Canada’s commitment to multiculturalism could help it become a world leader

U of T physicists discovered a way to increase the resolution of microscopes and telescopes

Team led by U of T researchers discovers energy source sustaining microbial life deep beneath Earth’s surface

Read about the latest findings in Deep Carbon Observatory

Read more about Barbara Sherwood Lollar's research

Read about Barbara Sherwood Lollar being honoured by the Royal Society of Canada

Making waves: How the �ؿ�ζSM made the discovery of gravitational waves possible

Didn't see that stop sign? U of T researchers examine the invisible world of human perception

We see much less of the world than we think we do

Objects play fundamental role in how we focus our attention

New element added to research: colour

U of T physicists taking a cold look at atomic interactions