Faces of the Foundation: Adam Marblestone

February 20, 2018
Bennett McIntosh
Livermore, Calif

“The brain is the most interesting thing that exists,” says Adam Marblestone – but the very intricacy that makes the brain so fascinating (and allows us to even have feelings like fascination) also make the brain incredibly difficult to model and understand scientifically. Marblestone has devoted his career to changing that.

“There’s a lot of technical virtuosity being applied in neuroscience, but the scale of the problem is still vastly beyond what we can do right now,” he said. “There are still million-fold improvements needed” in the cost and efficiency of the tools measuring the brain for the instruments of science to come close to comprehending it in its entirety.

Such rapid advances are not unprecedented – at the twilight of the $3 Billion Human Genome Project, it cost hundreds of millions of dollars to sequence a single human genome, but incomprehensibly rapid advances in the last decade have brought that cost down to a few thousand dollars. Now, Marblestone hopes to drive a similar revolution in neuroscience.

For his PhD thesis, frustrated by the vast gap between the complexity of the brain and the tools to study it, Marblestone went back to the figurative drawing board, examining what the absolute limit of neuroscience technology might be, and how to get there. Starting from only the hardest constraints in energy, volume, and data – “your technology can’t fry the brain, and it can’t pull apart the brain,” he quips – Marblestone drew on concepts from optics to information theory to molecular biology to lay what might be possible in neuroscience’s future.

“We really tapped into the Hertz network,” he recalls – he’s now co-authored papers, posters, and patents on innovative neuroscience tools with nine different Hertz fellows. Some tools, like a method combining revolutionary techniques in DNA sequencing and expansion microscopy, are now patented and being applied to large-scale brain mapping projects such as one being led by the US intelligence community’s advanced research agency IARPA, while others, like a still-theoretical technique for sensing neural impulses with fiber optic photonics, he regards more as “straw-man concepts” for experts in various fields to improve on.

Marblestone now splits his time between academia and industry, serving as Chief Strategy Officer of Kernel (a neural interfacing hardware company aiming to apply its technologies to the human brain) and a co-founder of BioBright (which creates tools for scientists to augment and automate their research process). At MIT, he works in the lab of Hertz Fellow Ed Boyden, working to develop the tools necessary to, say, find all the millions of neural connections in one cubic millimeter of brain.

“What’s interesting about academia,” he says, “is this constant rethinking you get through the students. You can have generations of improvement in the space of a few years”

For the past four years, Marblestone has helped select the Hertz Fellows who will do that rethinking. By interviewing candidates for the Hertz Fellowship, he says “You know that you’re not interviewing them to keep doing the same thing that they’re doing. We’re selecting candidates who will surprise us.”