Gunnar Newquist

Gunnar Newquist

Founder and CEO

Dr. Gunnar Newquist is a scientist turned entrepreneur obsessed with the functionality of the brain and how living creatures learn. He obtained a Ph.D. in Cell & Molecular Neuroscience from the University of Nevada, Reno, and conducted a post-doctorate in learning and memory at ESPCI Paris-Tech. Gunnar is also a trained concert pianist and a former professional extreme skier.

Origin of Brain2Bot

During my PhD, I became intensely interested in learning and behavior. I had been studying the molecular development of the nervous system, and we had a lot of information about the structure of the brain, but there was little connection between this fundamental structure and how that structure resulted in the emergent property of behavior. The more I looked into it, the more I realized that the principles of molecular neuroscience weren’t actually being considered when creating theories of behavior. Instead, ideas were being carried over from a pre-scientific era. Some of these ideas were completely untestable (a hallmark of pseudoscience), and some ideas were even directly contradicted by hosts of controlled experiment. This impregnation of sub-standard science cut to the heart of who I am. My field was allowing tradition to override proper theory. I became more and more frustrated with the state of affairs because the experiments that I knew had to be conducted to understand how the brain really works were not the kind of experiments being funded. If I continued on that academic track, I would be spending my days arguing, rather than progressing. For example, one of my articles went through no less than 20 reviewers and took 8 years to publish! That’s not an efficient use of time for anyone.

 

However, an alternative pathway presented itself. I had been following the progression of artificial intelligence, and especially, the invention of self-driving cars. Some of the engineering of this technology was directly inspired by neuroscience—the kind of neuroscience I had been at odds with. Through the AI advances, I saw many predictable failures that stemmed from the failures in neuroscience theory. I knew I could do better through the application of testable, bottom-up neuroscience. After all, even small insects like bees can navigate in amazingly complex and dynamic environments. By using simple models, rather than humans as inspiration, I could build the system step-by-step from the well-established principles. For the things I didn’t know yet, I could design the right experiments and figure out the fundamentals of how it worked. In the end, if I was right, the artificial intelligence, and the autonomous robots, would work better and be more life-like. No more guessing—just direct answers.

 

So that is why I started Brain2Bot: to first understand the brain, and to second, build smarter robots. The best part is, the better I am at one, the better I will be at the other. Now, there is no argument. If the theory doesn’t hold up under a test, we abandon it. If the software doesn’t work well enough, we try something else. My small team has already created the world’s most efficient learning algorithm from this core tenant. We have made very large advances starting from scratch, and the most enticing prospects are just on the horizon. I think we can have it all—proper scientific discovery married to technological breakthrough in one place. This is just the beginning of something very exciting for us, and it all comes from a love of science.