Carney Institute (CI): Tell us a bit about yourself.
Arif Hamid (AH): I grew up in Ethiopia and moved to the U.S. 20 years ago. I did my undergraduate work at the University of Minnesota, my graduate work at the University of Michigan, Ann Arbor, and my postdoctoral work at Brown. I returned to the University of Minnesota to start my own research group in the Neuroscience Department.
CI: Did you have an interest in science from an early age?
AH: I have always loved math, physics, and the hard sciences. I started my undergraduate studies in aerospace engineering with the goal of being an astronaut. On a trip to the NASA, Johnson Space Center in Houston, I received life-changing advice from one of the staff engineers, and he put things into perspective. He shared that most aerospace engineering technologies were being developed in or for the military, and, at the same time, the civilian aviation industry was stagnating, and NASA’s budget was getting cut year after year. So, I started to re-examine my career trajectory. In retrospect, I was very interested in exploring the mechanics of why my own hyperactivity was uncontrollable, with gifts of forgetfulness, hyperfocus modes, and severe procrastination. I now study exactly this; brain dopamine mechanisms that regulate processes, including motivation, learning, and planning. In a profoundly fulfilling full circle, I published a collaborative project with NASA last year on unique AI solutions to biomonitoring, space health, and space science during deep space travel [Article 1, Article 2].
CI: Tell us about your postdoc work at Brown.
AH: At Brown, I worked with (Carney Institute Associate Director) Christopher Moore and (Center for Computational Brain Science Director) Michael Frank. Michael is adjacent to my field in dopamine and decision-making, whereas Chris is an expert in neocortic dynamics, somatosensation, and computational functions of neuro-vascular interactions. So, it was a rich ecosystem of research interests that collided nicely.
Moreover, as a postdoc, things fail a lot. You're just past that learning stage in graduate school, and every action or effort costs you time, money, and the not-so-infrequent sweat and tears. But, both Chris and Michael believe in their trainees’ success on a much longer timescale, not just as it relates to an individual experimental outcome. So, they were the ideal mentors at that time in my career. They also did not have any problem relinquishing a level of control and autonomy to me; this profoundly accelerated my on-the-job training in adult-scientist duties.
CI: Did you find a lot of cross-pollination at Carney?
AH: Absolutely. I spent a lot of time in the Innovation Zone and Computation Lab at Carney. Bringing my laptop, preparing a talk, or doing some analysis there allowed me to bump into a variety of people from CLPS, Engineering, and some of (Center for Computational Brain Science Associate Director) Thomas Serre’s students who were doing vision-based computation. There was also a representation of people who had experimental expertise but were hanging out in this computational space to sharpen how they think and talk about their projects and get feedback from their peers. This is an immensely undervalued opportunity, and I was lucky to benefit from this investment from the Carney Institute.
CI: Tell us more about your research that focuses on the formalization of dopamine's role in reinforcement learning.
AH: When we think about performing an action or a task, we have to engage in a series of operations. For example, if there’s some water in a cup in front of me, I first have to want it or be compelled to pursue it based on how thirsty I am. This is a critical motivator or instigator process; its job is to kindle additional operations, including setting a goal (or plan) to reach for the cup. As I'm reaching for it, I must monitor how well I'm doing; is my arm moving in the right direction? Am I grasping the cup sufficiently and bringing it to my mouth, etc? Finally, when I sip the water, it will hopefully quench my thirst – It will be delicious. So, this final stage of attaining a reward creates a sort of ‘thank-you note’ for a job well done – the technical term for this is a reward credit. This reward credit is paid to the collection of brain circuits that helped achieve this feat, from initially planning the reach to monitoring and executing it.
By the way, while the example I provided above is centered on moving, our internal thoughts have exactly the same mechanics. So, there is a deep analogy between cognition and movement.
The processes I described – desires, compulsions, plans, error/progress monitoring and reward-credit assignment are all regulated by a critical neurotransmitter in the brain – dopamine! This link, and dopamine’s royalty status in cognition, has been made and strengthened by observations that psychiatric or neurological disorders with symptomology reducible to the listed processes (lack of motivation or vigor in depression and Parkingson’s, intense compulsion to gamble, inappropriate error monitoring, or reward credit learning in addiction) are due to dopamine dysregulation. Almost all current therapies aim to restore normal dopamine functions.
So, over the last 15 years or so, I have been studying which biological components in the brain help to perform these cognitive processes. We have made many experimental discoveries and provided conceptual breakthroughs connecting previous gaps in understanding. For example, my doctoral studies showed that dopamine levels fluctuate in a deep brain structure called the nucleus accumbes, and that these changes cause animals to engage with the environment vigorously. This helped to clarify how learning and motivation are connected. My postdoctoral work at Brown discovered that dopamine pulses travel across brain areas in a particular rhythm–called dopamine waves and that carefully orchestrated dopamine waves regulate adaptive behaviors. My lab at Minnesota is now investigating how this pattern is created and how disrupting this pattern produces signatures of psychiatric disorders.
CI: Given that context, is dopamine doping a real possibility, or is it science fiction?
AH: Interesting question! I am not sure, but I would say that it’s possible. However, rigorous technical, experimental, or theoretical aspects from our field haven't focused on testing these ideas. From a purely behavioral point of view, I can tell you that anticipating a future outcome and engaging in planning its pursuit is the most potent link to dopamine measured in the brain. The dopamine doping strategies you mention here are likely attempting to elevate dopamine levels to help plan and focus in the first place, so there is a chicken-or-egg conundrum. My intuition is that controlling other factors associated with achieving hard cognitive tasks (like the desk I always write my papers from and the time of day and music I use to enter the flow state) can help to provide the brain chemistry needed to get things done. This intuition is based on empirical studies we did that suggest that “dopamine signals the value of work,” so in an environment where my work is valuable, I would predict dopamine would track. But I am not in the business of giving advice about dopamine doping because I honestly don’t know.
CI: Can the genes that produce dopamine be turned on and off by behavioral or environmental stimuli?
AH: This is not an area that I actively study, but it’s one that I’m very much interested in, especially in the context of reinforcement learning. For example, there's literature that shows that if you have animals that were raised cross-generationally with fragmented maternal care and in stressful environments, also known as an early-life stress model, those animals will develop a hyperactive dopamine circuit such that they are a little bit more exploratory and prone to substance abuse. There is some evidence that this is due to changes in dopamine-related genes during development that can shape behaviors emitted at different developmental changes. This is an area of intense investigation, with many unanswered questions.
Another example I can think of is studies showing that female animals show different dopamine activity levels that are attuned to hormonal cycles and decision motivators like food, caregiving, or mating. Yet, we know very little about the circuit, computational, endocrine, and behavioral mechanisms.
I think the challenge (and opportunity) for systems and computational neuroscience is to frame examples like these — inherently spanning genetic, cellular, neuroendocrine and behavioral levels of analysis — into computations and mathematical formulations that allow for specifications of normative, generalizable and falsifiable hypotheses about these processes that evolved to meet the demands of an ever-changing world.
CI: Where do you see the field going next?
AH: There are two areas that I'm passionate about. One is in the science domain, and the other is in the mentoring/outreach domain.
Some of the studies I did at Brown, and now in my own lab, on dopamine waves are significantly changing how the field thinks about how it shapes the underlying circuits and how it breaks down in psychopathology. This has helped to reinvigorate an older family of ideas that are starting to interface quite seamlessly with disorders in psychiatry and the architectures of modern artificial intelligence systems. There are significant opportunities to leverage our understanding of how the brain works to alleviate human suffering and also to inform the next generation of biologically inspired AI, which has diverged from design principles in natural intelligence in recent years.
In the mentoring realm, I'm very passionate about training the current generation of junior scholars into deep thinkers and thought leaders within my field. I also care deeply about supporting healthy academic infrastructures that facilitate an environment that allows many more future generations of thinkers to flourish.
This requires, on the one hand, having mentors like Chris and Michael who are committed to the long-term successes of their trainees and have a deeply optimistic bias towards their success. In this spirit, I now leverage my position as a university professor to build environments of inclusion and excellence within my research group and university, which is focused on trainee development and team building. It's not easy, but I am starting to get a hang of it.
The second prong has to be outreach! I am engaged in global outreach in research capacity building on the continent of Africa, including my home country of Ethiopia. The continent is rapidly developing, and a large percentage of the population is younger than 25. So, there is a wonderful opportunity to tap this potential through scholarly engagement and capacity building that would empower African scholars to tackle fundamental challenges in their communities. I co-direct the Imbizo Computational Neuroscience summer course in Cape Town, South Africa, in addition to formal collaborations and partnerships with universities in Ethiopia. The goal is to provide an alternative to existing training opportunities that require African scholars to travel abroad with little incentive or promise to return, causing a significant brain drain problem– I am a perfect example of this! On the continent, there are problems that Africans see every day. I think that if you empower people on the ground - train them at the forefront of science and give them the tools to solve any given problem - then they wouldn't have to rely on foreign aid to solve their problems. So, I am convinced that helping to build academic infrastructure on the continent will yield long-term, non-linear returns. I hope you can appreciate that this is a deeply fulfilling endeavor for me!