The Short Version

Brain as blueprint: Morgan Sammons on what computers can learn from evolution

Episode Summary

Morgan Sammons, associate dean for natural sciences and mathematics in UAlbany's College of Arts and Sciences, takes us behind the scenes as neuroscientists, mathematicians, physicists and nanoscale engineers at UAlbany are working together to advance to the field of neuromorphic computing and design the next generation of powerful computers inspired by the human brain.

Episode Notes

The longer version:

 

Morgan spoke to us as a representative of a larger cluster of scientists and scholars at UAlbany thinking about how the next generation of computers can take inspiration from the human brain. 

His work with that team of biologists, psychologists, mathematicians and nanoscale engineers underscores a truth many scientists have long understood: the problems we care most about, such as how to live long and age well, will not be solved by researchers in one discipline alone. 

Neuromorphic computing is a prime example.

“These fields are really starting to blur together in ways that certainly, when I started my career 25 years ago, wasn’t the case,” Sammons said. “At the fundamental level many scientific fields are remarkably similar. And there’s certainly a group of people who say, ‘Well, every field is just math when it comes back to it.’ 

But when we think about the people designing computer chips, or we think about people who are doing sustainable engineering, or mechanical or electrical engineering, or nanotechnology, many of them know biology. Many of them have done biology in the past. So often they’re drawn toward problems that are biology or life sciences- related.”

Part of that, he said, is innate human interest in the problems that affect us personally. 

“There’s a huge push on campus for healthy aging — the idea that getting older is something that all humans must confront. So you need to bring all these disciplines together. Whether or not there are similarities between what I do and what a computer engineer might do, or a computer scientist might do, we all come together on the fundamental societal problems. That’s why things like UAlbany’s AI Plus Initiative are so successful. Because everybody touches it, and it will impact everything we do. 

It’s not just the scientists doing drug discovery or trying to understand DNA sequences. It’s also the English professors analyzing hundreds of millions of works over the span of time and understanding grammar and structure; it's our philosophers sitting down and thinking about, ‘How does this impact our society?' Whether it’s just AI generally or the more specific part that we’re talking about — creating new computer chips, which is pretty niche — AI is going to touch a lot of people’s work.” 

Go deeper

Learn more about Morgan's day job as part of UAlbany's RNA Institute studying, as he puts it, "what it is about our DNA that makes us who we are."

Recently, that work included a collaboration with fellow RNA scientists to study how our cells respond to and fight off diseases like Zika virus — and how that might help develop better treatments in the future.  

Visit the Sammons Lab

Morgan also mentioned UAlbany’s AI Plus Initiative and Center for Healthy Aging

Campus news

Upcoming events

Explore everything happening on campus with the University at Albany Events Calendar

Episode credits

Audio editing and production by Scott Freedman
Headlines by Erin Frick
Hosted and written by Jordan Carleo-Evangelist

Episode Transcription

0:01 Jordan Carleo-Evangelist

Welcome to The Short Version, the UAlbany podcast that tackles big ideas, big questions, and big news in less time than it takes to cross the Academic Podium. I'm Jordan Carleo-Evangelist in UAlbany's Office of Communications and Marketing. 

In the Emmy-winning television series The Bear, a character defined by his unsentimental number crunching is known simply as computer. 

It's funny, but there's truth there. Despite the dizzying pace of advances in artificial intelligence in quantum computing, the most powerful and mysterious computer known to science is still the human brain. 

Understanding how a seemingly impossibly complex network of neurons and synapses makes humans such nimble problem solvers is key to building the next generation of computers — computers that can creatively tackle challenges they've never encountered before and use less energy while doing it. That means we need more than hardware and software engineers. We also need experts in biology, psychology, and neuroscience.

That's where Morgan Sammons enters the conversation. 

Morgan is an associate professor in UAlbany's Department of Biological Sciences whose work has largely focused on how the information in our DNA makes us who we are. 

But Morgan also serves as associate dean for natural sciences and mathematics in UAlbany's College of Arts and Sciences, and that job increasingly puts him in the middle of conversations among scientists like him who study living systems and those building the artificial intelligence systems of the future. 

We sat down with Morgan to talk about why many believe the future of AI and quantum computing are deeply entangled with the study of biological systems and fields like quantitative neuroscience — and why the secrets of our own brains have been so difficult to discover.

1:53 Morgan Sammons

Biology and biological systems, whether it's a very simple decision made by a bacterium or whether it's a brain and it's an organism making a decision, that's the result of hundreds of millions of years of trial and error through evolution. Whereas if we go to something like a computer chip, which is so instrumental for AI and for sort of the future of a lot of what we're doing, modern computer chips have only been around for 80 years, so we're trying to take inspiration from biology and from biological systems to sort of use that as an inspiration to design the next generation of as many things as we can. 

I'm a geneticist who studies DNA and who tries to understand why the information that we receive from our parents leads us to who we are and what we do. It turns out that that's a pretty interdisciplinary field in that we interface daily with computers and we need people from all disciplines really to solve the challenges both of today and the challenges of tomorrow.

The human brain is remarkable compared to computer chips. Right now, the current computer chips do things that our brains don't do that well, right? We can do arithmetic, but computer chips do that extremely quickly, extremely efficiently and quicker than most of us can do. But the human brain in particular can do things that right now the current computer hardware really can't even imagine. The brain can do lots of things simultaneously, so it can control vision, it can control language and decision-making and communication. It can think and it can create, and it can learn and adapt. Those are the types of activities that we want computer chips to be able to do.

There's a few reasons why the brain is very good at those things, one of which is that the brain is incredibly energy efficient. So imagine all of the things the brain has to do every single day, every single moment, and it can do all of that on the energy equivalent of about two cans of Coke.

You go online and you jump on chat GPT, and you do one simple prompt and you just ask it one question that requires enough energy to essentially run the human brain for 10 minutes. Now imagine the scale at which people are going online with current computer chips, how much energy that requires and how much of a carbon footprint that produces. We're told all the time to bring AI into the classroom, into our lives, into our companies, and as we do that, of course, it creates amazing efficiencies in some areas, but there are now inefficiencies within the power grid, within how we use and create energy, and the hope of using neuroscience to develop and create new computer chips is that electrical power savings, that efficiency. There's some really amazing potential to these approaches. 

The current limits of regular computer chips is size where we think about our fabrication plants here at UAlbany with NY Creates, and it's always about making things smaller and that makes it more efficient, but the smallest thing we know of are these molecules called qubits.

You can't go smaller, so when we're talking about making computer chips and having things being done in parallel and energy efficiently in a theoretical way, quantum computing is the best we can currently imagine in terms of efficiency and speed and ability. What quantum computing is going to do that is sort of even, in theory, better than this idea of brain-inspired neuromorphic computers or architectures is that it doesn't necessarily need the same type of architecture and power. It's using not electrical signals, but it's using fundamental properties of atoms and molecules. We think about logic, it's a yes or no. Well, molecules at the atomic level are in multiple configurations. Either it's spinning one way or another, and so can we take advantage of this fundamental, tiny, tiny little particle and use that to inform how computers work. But we're not quite where we need to be to make that a reality where it's a consumer product or anything other than in a big national lab.

Neuromorphic computing still uses traditional computer chips and the concepts that we know, but is sort of using that inspiration of the brain and what it does remarkably well, both that energy efficiency, but also doing things in parallel. We're pretty good at doing lots of things simultaneously. We can walk and chew bubble gum. Why neuromorphic computing is kind of a great way to go for now is because the brain's incredible at multitasking, and the brain is really great at adapting so it can learn, it can experience, and it can make changes over time, whereas a computer chip is going to require, at least in the architecture, can't really change. We can ask it to do different things, but we can't ask it to adapt on its own. We can't ask it to experience the world and then change itself, change where those connections are made, whereas the brain gets to do that every day.

Computer chips and computer systems — they're not naturally resilient. There are really critical failure points where a computer chip just shuts off. The brain is remarkably resilient when it comes to its ability to overcome issues day to day. Just imagine every time you stay up a little too late and you wake up, your brain doesn't stop having you breathe. Your brain can still think maybe it's a little less clear than normal, but those things, the important tasks, still work. The brain can adapt and can be resilient in ways that if you removed half a computer chip, you would not have quite the same resiliency. So that's something that we also really want to look to, which is how's the brain so good at fault tolerance or so good at being resilient? Even though the world is crazy and whether we get sick, whether we stay up late, we can still kind of do what we need to do. Right now, the brain still wins. And where individual people and groups of people sort of come out on top versus these supposed AI systems that are intelligent, which is the human brain is creative, it can invent, it can imagine, it can think, and it can do things right now that we say computer chips and systems can do, but it can't quite get certain things right because it's basing it off of things it's seen before.

Whereas we as humans with our brain and how it works over millions of years of evolution, it can create and it can imagine. And so I would say until AI systems are not just good at math and not just good at coding and some of these other problem solving that it has to do based off of examples that we've given it, that's really where we still have the advantage, I'd say. And hopefully no one's out there hoping for Skynet. That's hopefully not where we're going to go, but certainly it's something that we need to bring everybody to the table to really think about — not just how we make it, but how we use it.

9:27 Jordan Carleo-Evangelist

That was Morgan Sammons of UAlbany's Department of Biological Sciences and associate dean for natural sciences and mathematics in the College of Arts and Sciences. 

To learn more about Morgan's experience working across the aisle with his colleagues in mathematics, computer science and engineering, be sure to check out The Longer Version in our show notes.

Before we go this week, my colleague Erin Frick will catch you up on the latest news from across campus.

9:53 Erin Frick

This week, IBM announced the expansion of its partnership with the UAlbany Center for Emerging Artificial Intelligence Systems with seven new joint research projects that will use IBM's latest Spyre Accelerators installed on campus earlier this month. This new IBM computing hardware is designed to perform powerful AI tasks more efficiently and consume less energy.

Academic affairs last week officially unveiled the new First-Generation Scholars Lounge as part of the campus wide celebration of National First Generation Week. The lounge located in Taconic 265 provides a dedicated study and quiet break space for first-gen students. More than a third of Great Danes are the first in their families to attend college. 

Christina Phillips has been named UAlbany's interim assistant vice President for facilities management, taking over responsibility for planning major campus construction projects and maintaining UAlbany's roughly 6.7 million square feet of buildings. Phillips joined the Office of Facilities Management in 2019 after overseeing operations for the State's Office of Cultural Education. 

Looking ahead, there are two opportunities on campus this week to explore your love for foreign languages and cultures.

On Thursday, Nov. 20, French speakers of all skill levels are welcome at La Pause Cafe in Humanities 290 from 4-5 p.m. And on Friday, Nov. 21, the Italian table welcomes those interested in Italian language and culture to Humanities 131 from 2:45-3:45 p.m. These are recurring events, so check the UAlbany Events Calendar for future dates. 

Also on Friday, Nov. 21, UAlbany's AI Plus Institute will celebrate Foundational AI Day with a series of talks and panels featuring students, researchers, and guest speakers. Discussions will explore the latest on the large, complex machine learning models that underpin many AI applications, with a special focus on how they can be used in medicine. That event runs from 9 a.m. to 5 p.m. in the Campus Center Boardroom. 

Lastly, don't forget that students have until 5 p.m. on Dec. 1 to submit artwork for the 2025 Holiday Card Contest. The top prize is a $1,000 scholarship. You'll find a link to the full Holiday Card Contest details and all these stories in the Today at UAlbany News Center, and a link to the full University Events Calendar in our show notes.

12:11 Jordan Carleo-Evangelist

The Short Version would not be possible without contributions from many people, including for this episode: Scott Freedman, who provided audio production and editing support from the UAlbany Digital Media Studio, deep inside the Podium tunnels. And a news update from Erin Frick. 

We'll be back next week with a special Monday Extra Short on how the School of Social Welfare is using professional actors and role-playing to prepare students for the challenging conversations they'll face in their jobs. 

I'm Jordan Carleo-Evangelist here at the University at Albany, and this has been The Short Version.