Carey K. Morewedge, Boston University

Algorithms are a staple of modern life. People rely on algorithmic recommendations to wade through deep catalogs and find the best movies, routes, information, products, people and investments. Because people train algorithms on their decisions – for example, algorithms that make recommendations on e-commerce and social media sites – algorithms learn and codify human biases.

Algorithmic recommendations exhibit bias toward popular choices and information that evokes outrage, such as partisan news. At a societal level, algorithmic biases perpetuate and amplify structural racial bias in the judicial system, gender bias in the people companies hire, and wealth inequality in urban development.

Algorithmic bias can also be used to reduce human bias. Algorithms can reveal hidden structural biases in organizations. In a paper published in the Proceedings of the National Academy of Science, my colleagues and I found that algorithmic bias can help people better recognize and correct biases in themselves.

The bias in the mirror

In nine experiments, Begum Celikitutan, Romain Cadario and I had research participants rate Uber drivers or Airbnb listings on their driving skill, trustworthiness or the likelihood that they would rent the listing. We gave participants relevant details, like the number of trips they'd driven, a description of the property, or a star rating. We also included an irrelevant biasing piece of information: a photograph revealed the age, gender and attractiveness of drivers, or a name that implied that listing hosts were white or Black.

After participants made their ratings, we showed them one of two ratings summaries: one showing their own ratings, or one showing the ratings of an algorithm that was trained on their ratings. We told participants about the biasing feature that might have influenced these ratings; for example, that Airbnb guests are less likely to rent from hosts with distinctly African American names. We then asked them to judge how much influence the bias had on the ratings in the summaries.

Whether participants assessed the biasing influence of race, age, gender or attractiveness, they saw more bias in ratings made by algorithms than themselves. This algorithmic mirror effect held whether participants judged the ratings of real algorithms or we showed participants their own ratings and deceptively told them that an algorithm made those ratings.

Participants saw more bias in the decisions of algorithms than in their own decisions, even when we gave participants a cash bonus if their bias judgments matched the judgments made by a different participant who saw the same decisions. The algorithmic mirror effect held even if participants were in the marginalized category – for example, by identifying as a woman or as Black.

Research participants were as able to see biases in algorithms trained on their own decisions as they were able to see biases in the decisions of other people. Also, participants were more likely to see the influence of racial bias in the decisions of algorithms than in their own decisions, but they were equally likely to see the influence of defensible features, like star ratings, on the decisions of algorithms and on their own decisions.

Bias blind spot

People see more of their biases in algorithms because the algorithms remove people's bias blind spots. It is easier to see biases in others' decisions than in your own because you use different evidence to evaluate them.

When examining your decisions for bias, you search for evidence of conscious bias – whether you thought about race, gender, age, status or other unwarranted features when deciding. You overlook and excuse bias in your decisions because you lack access to the associative machinery that drives your intuitive judgments, where bias often plays out. You might think, “I didn't think of their race or gender when I hired them. I hired them on merit alone.”

When examining others' decisions for bias, you lack access to the processes they used to make the decisions. So you examine their decisions for bias, where bias is evident and harder to excuse. You might see, for example, that they only hired white men.

Algorithms remove the bias blind spot because you see algorithms more like you see other people than yourself. The decision-making processes of algorithms are a black box, similar to how other people's thoughts are inaccessible to you.

Participants in our study who were most likely to demonstrate the bias blind spot were most likely to see more bias in the decisions of algorithms than in their own decisions.

People also externalize bias in algorithms. Seeing bias in algorithms is less threatening than seeing bias in yourself, even when algorithms are trained on your choices. People put the blame on algorithms. Algorithms are trained on human decisions, yet people call the reflected bias “algorithmic bias.”

Corrective lens

Our experiments show that people are also more likely to correct their biases when they are reflected in algorithms. In a final experiment, we gave participants a chance to correct the ratings they evaluated. We showed each participant their own ratings, which we attributed either to the participant or to an algorithm trained on their decisions.

Participants were more likely to correct the ratings when they were attributed to an algorithm because they believed the ratings were more biased. As a result, the final corrected ratings were less biased when they were attributed to an algorithm.

Algorithmic biases that have pernicious effects have been well documented. Our findings show that algorithmic bias can be leveraged for good. The first step to correct bias is to recognize its influence and direction. As mirrors revealing our biases, algorithms may improve our decision-making.

Carey K. Morewedge, Professor of Marketing and Everett W. Lord Distinguished Faculty Scholar, Boston University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Brett Volmert, Michigan State University; Aitor Aguirre, Michigan State University, and Aleksandra Kostina, Michigan State University

How did your heart form? What triggered your first heartbeat? To this day, the mechanisms of human heart development remain elusive.

Researchers know the heart is the first organ to fully function in the growing human embryo. It begins as a simple tube that starts to pump blood by the fourth week of gestation. By the ninth week, the heart is fully formed. The heart is critical to early development because it provides essential nutrients throughout the developing fetus.

But due to its early formation, the heart is exposed for a long duration to substances a pregnant person might come into contact with, such as medications or pollutants. This may be a main reason why congenital heart disease is the most common type of birth defect in people, occurring in over 1 in 100 births worldwide.

Traditionally, scientists have used animal and cell models to study heart development and disease. However, researchers haven't been able to produce a cure for congenital heart disease in part because these models are unable to capture the complexity of the human heart. Due to ethical limitations, using human embryos for these studies is out of the question.

To help researchers study heart development and complications in pregnancy, our team of biomedical engineers and cardiovascular scientists have spent the past several years trying to create the next best thing: mini human hearts in the lab.

Human heart organoids

Organoids are complex 3D cellular structures that replicate significant aspects of the structure and function of a specific organ in your body. While organoids are not completely synthetic, functioning organs (yet), they still possess immense power to mimic key aspects of physiology and disease in the lab.

We created our heart organoids using a type of cell called a pluripotent stem cell. Although using these cells in research used to be controversial because they were originally derived from human embryos, this is no longer a concern, as they can be produced from any adult. Pluripotent stem cells have the potential to become any type of cell in the body. This means that cells from nearly any part of your body – typically blood or skin cells – can be turned into your own stem cells to grow your own mini heart.

By manipulating the ability of pluripotent stem cells to become any type of cell in the body, we guided these cells to become heart cells. The cells were able to self-assemble, replicating the main stages of human heart development during pregnancy. Our heart organoids have blood vessels and all the cell types found in the human heart, such as cardiomyocytes and pacemaker cells, which give them an edge over 2D cellular models. Furthermore, the electrophysiology and bioenergetics of these heart organoids are very similar to human embryonic hearts in ways that animal models aren't.

Our heart organoids beat like a tiny baby's heart, all while smaller than a grain of rice.

Pregnancy and the fetal heart

One area we're exploring with our heart organoids is maternal and fetal cardiac health. Maternal factors such as diabetes, hypertension or even depression can increase the risk of heart disease in newborns. Studying conditions that increase the risk of congenital heart disease can prevent and reduce the incidence of cardiovascular diseases worldwide.

We can mimic these maternal environments and simulate how they influence fetal heart development with heart organoids. For example, we used heart organoids to show that diabetes, a very common condition, increases the risk of heart disease in embryos. Compared to heart organoids created in healthy conditions, mini hearts exposed to diabetic conditions developed heart abnormalities like those of human fetuses and newborns with diabetic cardiomyopathy.

Our study found that diabetes-related developmental abnormalities of the heart are likely caused by an imbalance of omega-3 fatty acids, the building blocks of cell membranes and signaling molecules. However, dietary supplementation of omega-3 fatty acids could partially restore this imbalance and prevent diabetes-induced congenital heart defects.

Drug safety during pregnancy

The drugs pregnant people take can have significant health effects on both the parent and the fetus. Medications approved for use during pregnancy are not always safe, since adequate testing is complicated. Ethical concerns limit working with biological material from people, so researchers are left with animal models that aren't able to replicate human physiology closely enough.

Testing medications on human heart organoids allows researchers to better explore and predict potential harmful effects during pregnancy. One example is ondansetron (Zofran), a drug commonly prescribed to prevent nausea and vomiting during pregnancy. Although it has been linked with an increased risk of congenital heart disease, whether it causes the disease hasn't been confirmed.

We showed that heart organoids exposed to ondansetron had disturbed development of ventricular cells and impaired function, similar to what's seen in newborns exposed to ondansetron. Our findings provide data that may help update clinical guidelines on the use of the drug.

Another example concerns the use of antidepressants during pregnancy, which is associated with an increased risk of congenital heart defects. Selective serotonin reuptake inhibitors, or SSRIs, the most prescribed antidepressants in pregnant people, work by increasing the availability of serotonin in the body. Serotonin is an important molecule in cardiac development. Maternal serotonin, along with antidepressants, readily pass to the embryo and alter serotonin levels in the developing heart.

In the future, we plan to expose heart organoids to antidepressants and study their effects on the incidence of congenital heart defects. The results of such research on human heart organoids may also inform recommendations for drug replacement or repurposing.

Heart organoids have the potential to help scientists more precisely study how the human heart forms and how it develops disease. In the realm of medical innovation, we believe human heart organoids grown from stem cells are the beating promise of a healthier future.

Brett Volmert, Ph.D. Candidate in Biomedical Engineering, Michigan State University; Aitor Aguirre, Associate Professor of Biomedical Engineering, Michigan State University, and Aleksandra Kostina, Postdoctoral Researcher in Quantitative Health Sciences and Engineering, Michigan State University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Mansour Solaimanian, Penn State

While on the road, you're probably thinking more about your destination than the pavement you're driving over. But building roads requires a host of engineering feats, from developing the right pavement materials to using heavy equipment to lay them down. The better they're built, the longer roads last and the fewer construction delays drivers have to endure.

I am an engineer who does research on materials used in roads. Scholars in my field are working to develop materials that can make roads stronger and last longer.

Road materials

So, what are roads really made of? The simple answer is that they are made of typical construction materials such as aggregates – soils and rocks – as well as asphalt binder and Portland cement, which act like glue to bond it all together.

Asphalt binder is refined from crude oil. From crude oil, refiners first extract gasoline, kerosene and oil, and what remains at the bottom becomes the asphalt. Portland cement is manufactured using several different ingredients, including limestone, sand, clay, silica and alumina.

Engineers compact the mixture of asphalt binder and aggregates together at an elevated temperature, about 300 degrees Fahrenheit (150 degrees Celsius), which glues the aggregates together into the final product, called asphalt concrete.

If they're using Portland cement rather than asphalt binder to glue the aggregates together, the engineers cure the mixture of the cement and aggregates with water through a process called hydration.

Hydration bonds the cement to the aggregates to make the product, called Portland cement concrete, stronger. With this process, there's no external heating involved.

Pavement structure

Asphalt concrete's pavement structure typically has three main layers: the base layer, the intermediate layer and the surface layer.

Engineers call the existing ground where the pavement goes the subgrade. On top of the subgrade goes a new layer of unbound soil and stone, where the aggregates aren't glued together. This is called the subbase, or unbound aggregate base.

The base layer can be either stones packed together without any binding agent or a combination of stone and asphalt binder.

Once road builders make the base, it is time to build the asphalt concrete layers: the base layer, the intermediate layer and the surface layer. All these layers contain the aggregates – the pieces of rock and sand – glued together with the asphalt binder in some way.

Engineers determine how many layers to build and how thick to make each layer by figuring out how much traffic will drive over the road. The more traffic, the thicker the pavement needs to be. For example, on interstate highways, the depth of the layers combined could be 20 inches (51 centimeters) or more.

Building a strong road

The road builders place the material on the road with an asphalt paving machine called a paver. An operator runs the paver, which takes the materials from a truck and places them on the road. After that, heavy-duty rollers compact it down, make it strong and get it ready for vehicles.

For a strong and durable road, engineers first pick the best subgrade, or place on top of which to build pavement. If the subgrade is too weak, the road might crack and fail – even if the pavement uses the best materials.

First, the road builders use rollers to pack the subgrade down. Once they've compacted the subgrade, they place the stone aggregates directly on top of the subgrade and compact them down. This aggregate base on the subgrade provides a sturdy foundation for the asphalt layers.

If the road builders do not use the right materials, or do not put them together correctly, or do not design the pavement structure for the expected traffic, then the road can crack, rut and fail.

Cracking occurs either at extremely low temperatures or from heavy trucks and buses repeatedly driving over the road. Rutting, which refers to noticeable impressions in the road's surface, occurs mostly during summer heat under heavy trucks or at road intersections.

Potholes are a big road problem you've probably seen before. They often show up in the spring after water trapped in the pavement freezes over winter and then melts in spring. This melting process weakens the road, making it more breakable. Then, when vehicles drive over it, they can create potholes.

Before the road gets built, the materials undergo testing in a laboratory to make sure they can stand the loads from traffic and environment.

Engineers in the lab expose the pavement materials to both freezing and very hot temperatures to make sure they can withstand any weather. They also expose the pavement materials to water to make sure the materials will not fall apart if it rains or floods.

At the Penn State pavement laboratory, my team is testing asphalt mixtures to which we've added substances called modifiers. These include special polymers and fibers that could make the road stronger.

The next time you're on the road, remember that it takes a good amount of engineering and tremendous teamwork to create that smooth pavement surface you drive on.

Mansour Solaimanian, Research Professor, Larson Pennsylvania Transportation Institute, Penn State

This article is republished from The Conversation under a Creative Commons license. Read the original article.