I grew up on Orcas Island in Washington State and was a founding member of the Orcas Island Rowing Team, eventually becoming the first member to row for a D1 team in college.
At Boston U I studied Mechanical Engineering but my true interest was in rowing and figuring out how to squeeze every last watt out of my 5’10” frame to keep up with 6’5” Olympian teammates. My research led to exercise physiology and I spent more time reading papers, blogs, and physiology chat boards than studying mechanics and thermodynamics.
One thing that always fascinated me was how much more effort I could give in racing than I could in training. In racing the pain was easier to deal with, and I could push myself right to the edge of passing out. In training I would struggle to “go to the well” and the same work rates (wattage) would feel harder. This couldn’t be explained by the cardiovascular, pulmonary or muscular systems I was reading about in that era’s exercise physiology texts. Something was missing, and this stuck with me.
After college I got a job as an aerospace manufacturing engineer (as most MechE’s do in the PNW). I switched from rowing to racing triathlons and continued reading about training, racing, nutrition, and general physiology. I spent far too much time reading Slowtwitch, a triathlon message board where the best exercise physiologists of the time would spar with each other on the latest research and ideas.
Eventually I came across a theory which, at the time was revolutionary, although now seems obvious - the Central Governor Theory. It states that the brain is ultimately in control of physical output, and that it uses a combination of inputs (physiological state, external context, etc) to decide how much output it will allow. Therefore your maximal physical output at any given time has a “governor” given a number of different variables. This makes sense - we can’t run ourselves to death (without something going wrong) and people cheering us on gives us a boost of energy. Again, this seems obvious, but at the time the dogma was that your plumbing (cardio, pulmonary, musculature) dictated maximal physical output in endurance activities.
The Central Governor Theory was posited by a somewhat controversial exercise physiologist named Timothy Noakes. He wrote what is probably one of the most comprehensive books on the physiology of running, Lore of Running. At 922 pages in text book form, it’s extensive.
Noakes and other prominent exercise physiologists would frequent Slowtwitch and get into arguments with each other. I would learn more from those threads than reading 20 scientific papers. Not just in facts about physiology, but also how to think critically and challenge ideas from multiple angles. My favorite debates were between Noakes and Andrew Coggan, the (debatable) inventor of Functional Threshold Power and Training Stress Score used to measure cycling training loads. You could literally watch a scientific field evolve in real time during these debates.
I was a pretty terrible employee for the same reasons I was a terrible student. I spent most of my time reading these threads and papers or planning my training, but was able to get my work done in mad dashes between 10am and 2pm on Fridays. Despite that I was still rated highly, garnering top raises and projects. This is more a testament to how dysfunctional the aerospace industry is.
I was also very unhappy. I hated spending 8 hours a day sitting at a desk feeling like I was spinning my wheels and wasting precious time. Doing great work felt actively discouraged. When I worked fast (or even at a normal pace) the project lead would tell me “ok, let’s slow down and I’ll review this” which would take days. I would eventually just buy the parts and build the thing without clearance or review, and then get back to reading papers and Slowtwitch threads.
One day after a triathlon in Las Vegas (Silverman for you aficionados), I got pretty hammered and spent the entire night talking about the Central Governor Theory. The person I was traveling with finally got fed up and said “why don’t you just go work on that???”. The light bulb went on.
I spent the next few months looking at how I could do that. Being an engineer, what interested me most was not that there was a central governor (at this point the research was pretty definitive), but what physical neural circuit mechanisms resulted in there being a central governor? I decided to retrain as a neuroscientist, starting from the beginning with biology, chemistry and the general prerequisites for a neuroscience PhD.
I quit my engineering job, enrolled in North Seattle Community College and started coaching triathletes and cyclists as part of Cycle University to pay for it. The coaching afforded me more time for studies and helped refine the questions I was most interested in. The schooling also demonstrated that the academic probation in undergrad (now 10 years prior) was not representative of what I was capable of. The change was dramatic, but I truly loved feeling like I was taking control of my life, and working full time on what had interested me for almost a decade.
Coaching was one of the most rewarding experiences of my life. I watched all my athletes move from middle of the pack to vying for world championship slots. The tri team grew from 6 to 60 members and the indoor cycling program another coach and I started grew into Cycle University’s flagship product. The programs grew because of something very simple that has become harder and harder to cultivate, particularly over the past few years post-Covid: community. I focused very hard on not just coaching, but also ensuring members knew they had a place and felt connected with each other. This formed the basis for how I manage teams.
In originally thinking about the Central Governor, I hadn’t put a lot of thought into community or culture. I knew people cheering helps a person perform better, but I hadn’t thought about how the day-to-day reinforcing support that a strong community of people, and a culture focused on being better every day aimed at a common goal can be a performance enhancer. This seems obvious now and there are many anecdotes of how a high performance group working over time can achieve some of the best athletic performances the world has ever seen (Kenyan culture of the 2000’s, the Brownlee brothers and Great Britain in the early 2010’s, Eliud Kipchoge’s training camps and partners, and most recently Norway’s 4 golds at the 2020 olympics).
Professional coaching is also the ultimate management crucible. People are paying you to make them better in a sport. They’re your client, but also an advertisement for how good you are. You need them to 1) perform really well and 2) be happy doing it, or they will leave. This forces you to learn how to get people to work hard and with excellence, but also enjoy it. I will write more on that in the future and I’ve taken these lessons with me everywhere I’ve gone.
After 3 years of prerequisites, coaching and working in a neuroscience lab at the University of Washington studying birdsong, I moved to San Francisco to work on a PhD at UCSF. I joined Anatol Kreitzer’s lab where they were studying how motivation is turned into movement - mainly in the context of Parkinson’s disease. It was a perfect fit, and I was even able to study how motivation centers in the brain control running output (Roseberry et al., 2016).
During this time I ran a lot as it was the easiest sport to integrate into a busy grad student schedule, did some cycling, and raced a couple swimrun races. I also started drifting away from the initial purpose, feeling like I had answered many of my initial questions and there was more I could contribute than getting people to run faster. It was also easy to be nerd sniped with all the world class research going on around me. Everything looked interesting and biotechnologies were undergoing a boom in SF.
I became especially interested in restoring movement to people who had lost it as I couldn’t imagine a life without movement. After graduation I joined a brain-machine-interface (BMI) lab in Berkeley with Jose Carmena looking at how neural implants could be used to control prosthetics and restore movement in paralyzed patients.
After 7 months the bureaucracy of academia wore me out, and I left academia to join Elon Musk’s Neuralink and have a more direct path to helping people. This Wait But Why article was a major inspiration. Neuralink was a major growing experience learning to work in teams again, how to move ridiculously quickly towards a goal, and that amazing science can be done in companies / startups. The coolest science I’ve done was there, and resulted in this patent.
After Neuralink I worked on a number of different projects in various startups: built a multi-photon 3D printer for cellular devices, studied gene + drug therapies (DREADD-like), created the largest epigenetic study in dogs, invented an in vivo platform that can predict clinical trial outcomes and figured out why women experience worse side effects on GLP-1s than men.
I’m optimistic about the future of humans and stoked to continue building new and interesting technologies that accelerate us into the future.