AI

Understanding our place in the universe | MIT News

5 Mins read


Brian Nord first fell in love with physics when he was a teenager growing up in Wisconsin. His high school physics program wasn’t exceptional, and he sometimes struggled to keep up with class material, but those difficulties did nothing to dampen his interest in the subject. In addition to the main curriculum, students were encouraged to independently study topics they found interesting, and Nord quickly developed a fascination with the cosmos. “A touchstone that I often come back to is space,” he says. “The mystery of traveling in it and seeing what’s at the edge.”

Nord was an avid reader of comic books, and astrophysics appealed to his desire to become a part of something bigger. “There always seemed to be something special about having this kinship with the universe around you,” he recalls. “I always thought it would be cool if I could have that deep connection to physics.”

Nord began to cultivate that connection as an undergraduate at The Johns Hopkins University. After graduating with a BA in physics, he went on to study at the University of Michigan, where he earned an MS and PhD in the same field. By this point, he was already thinking big, but he wanted to think even bigger. This desire for a more comprehensive understanding of the universe led him away from astrophysics and toward the more expansive field of cosmology. “Cosmology deals with the whole kit and caboodle, the whole shebang,” he explains. “Our biggest questions are about the origin and the fate of the universe.”

Dark mysteries

Nord was particularly interested in parts of the universe that can’t be observed through traditional means. Evidence suggests that dark matter makes up the majority of mass in the universe and provides most of its gravity, but its nature largely remains in the realm of hypothesis and speculation. It doesn’t absorb, reflect, or emit any type of electromagnetic radiation, which makes it nearly impossible for scientists to detect. But while dark matter provides gravity to pull the universe together, an equally mysterious force — dark energy — is pulling it apart. “We know even less about dark energy than we do about dark matter,” Nord explains.

For the past 15 years, Nord has been attempting to close that gap in our knowledge. Part of his work focuses on the statistical modeling of galaxy clusters and their ability to distort and magnify light as it travels through the cosmos. This effect, which is known as strong gravitational lensing, is a useful tool for detecting the influence of dark matter on gravity and for measuring how dark energy affects the expansion rate of the universe.

After earning his PhD, Nord remained at the University of Michigan to continue his research as part of a postdoctoral fellowship. He currently holds a position at the Fermi National Accelerator Laboratory and is a senior member of the Kavli Institute for Cosmological Physics at the University of Chicago. He continues to investigate questions about the origin and destiny of the universe, but his more recent work has also focused on improving the ways in which we make scientific discoveries.

AI powerup

When it comes to addressing big questions about the nature of the cosmos, Nord has consistently run into one major problem: although his mastery of physics can sometimes make him feel like a superhero, he’s only human, and humans aren’t perfect. They make mistakes, adapt slowly to new information, and take a long time to get things done.

The solution, Nord argues, is to go beyond the human, into the realm of algorithms and models. As part of Fermilab’s Artificial Intelligence Project, he spends his days teaching machines how to analyze cosmological data, a task for which they are better suited than most human scientists. “Artificial intelligence can give us models that are more flexible than what we can create ourselves with pen and paper,” Nord explains. “In a lot of cases, it does better than humans do.”

Nord is continuing this research at MIT as part of the Martin Luther King Jr. (MLK) Visiting Scholars and Professors Program. Earlier this year, he joined the Laboratory for Nuclear Science (LNS), with Jesse Thaler in the Department of Physics and Center for Theoretical Physics (CTP) as his faculty host. Thaler is the director of the National Science Foundation’s Institute for Artificial Intelligence and Fundamental Interactions (IAIFI). Since arriving on campus, Nord has focused his efforts on exploring the potential of AI to design new scientific experiments and instruments. These processes ordinarily take an enormous amount of time, he explains, but AI could rapidly accelerate them. “Could we design the next particle collider or the next telescope in less than five years, instead of 30?” he wonders.

But if Nord has learned anything from the comics of his youth, it is that with great power comes great responsibility. AI is an incredible scientific asset, but it can also be used for more nefarious purposes. The same computer algorithms that could build the next particle collider also underlie things like facial recognition software and the risk assessment tools that inform sentencing decisions in criminal court. Many of these algorithms are deeply biased against people of color. “It’s a double-edged sword,” Nord explains. “Because if [AI] works better for science, it works better for facial recognition. So, I’m working against myself.”

Culture change superpowers

In recent years, Nord has attempted to develop methods to make the application of AI more ethical, and his work has focused on the broad intersections between ethics, justice, and scientific discovery. His efforts to combat racism in STEM have established him as a leader in the movement to address inequities and oppression in academic and research environments. In June of 2020, he collaborated with members of Particles for Justice — a group that boasts MIT professors Daniel Harlow and Tracy Slatyer, as well as former MLK Visiting Scholar and CTP researcher Chanda Prescod-Weinstein — to create the academic Strike for Black Lives. The strike, which emerged as a response to the police killings of George Floyd, Breonna Taylor, and many others, called on the academic community to take a stand against anti-Black racism.

Nord is also the co-author of Black Light, a curriculum for learning about Black experiences, and the co-founder of Change Now, which produced a list of calls for action to make a more just laboratory environment at Fermilab. As the co-founder of Deep Skies, he also strives to foster justice-oriented research communities free of traditional hierarchies and oppressive power structures. “The basic idea is just humanity over productivity,” he explains.

This work has led Nord to reconsider what motivated him to pursue a career in physics in the first place. When he first discovered his passion for the subject as a teenager, he knew he wanted to use physics to help people, but he wasn’t sure how. “I was thinking I’d make some technology that will save lives, and I still hope to do that,” he says. “But I think maybe more of my direct impact, at least in this stage of my career, is in trying to change the culture.”

Physics may not have granted Nord flight or X-ray vision — not yet, at least. But over the course of his long career, he has discovered a more substantial power. “If I can understand the universe,” he says, “maybe that will help me understand myself and my place in the world and our place as humanity.”


Source link

Related posts
AI

Researchers at Stanford Use AI and Spatial Transcriptomics to Discover What Makes Some Cells Age Faster/Slower in the Brain

3 Mins read
Aging is linked to a significant rise in neurodegenerative diseases like Alzheimer’s and cognitive decline. While brain aging involves complex molecular and…
AI

This AI Paper from Anthropic and Redwood Research Reveals the First Empirical Evidence of Alignment Faking in LLMs Without Explicit Training

4 Mins read
AI alignment ensures that AI systems consistently act according to human values and intentions. This involves addressing the complex challenges of increasingly…
AI

TOMG-Bench: Text-based Open Molecule Generation Benchmark

3 Mins read
Molecule discovery is important in various scientific research fields, particularly pharmaceuticals and materials science. While the emergence of Graph Neural Networks (GNNs)…

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *