This is the full interview for the story - “Science Corp. Aims to Plant Ideas in Brains with New Device.” The conservation took place on January 10, 2025 at Science Corp.’s headquarters in Alameda, California.
These transcripts are part of our effort to give readers the source material behind stories and a chance to go deeper on the science and technology for those who want it. This will not be for everyone but should please others. I’ll mark the transcripts with the “context window” subhead for easy spotting on the homepage until I change my mind.
The discussion included the following individuals:
Ashlee Vance - Reporter at Core Memory
Max Hodak - CEO of Science Corp.
Alan Mardinly - Director of Biology at Science Corp.
Yifan Kong - Director of Microfabrication at Science Corp.
Hodak: You’re familiar with the basic overall high-concept pitch of neural engineering. There's, of course, no one approach. There are all kinds of different ways you can read and record information from the brain, but they all have trade-offs.
You could place electrodes into the brain and get more bandwidth. You could place electrodes in the blood vessels and get less bandwidth. You could do things in between. You could use ultrasound, but that requires genetic modifications to get very high functional resolution. All of these things have complex trade-offs. And so the corollary is that the appropriate probes depend on the application.
You’re not going to come out with the BCI. You're going to come out with a motor prosthesis, or a visual prosthesis, or an auditory prosthesis. Or you're going to do modulation for epilepsy or depression. And obviously, there's a lot of progress that's been made on these things, which I'm all for. But I contend that this is not the promise of BCIs. These are intermediate things that will get done along the way. The promise of BCIs is getting big cognitive objects out of the brain, and redrawing the borders around the brain.
We want something that is capable of simultaneous stimulation and recording over large areas of the brain, at single-cell resolution. Ideally with cell-type specificities because there are so many different types of neurons, and if you have an electrode in there, you can't tell a glutamatergic neuron apart from a dopaminergic neuron, or any of these things. And we now do have tools in neuroscience that allow us to record larger numbers of cells, or study the larger number of cells, but that will require genetically modifying the brain. There are pros and cons to these different things, but none of the known conventional neural interface technologies meet these requirements, even in principle.
Keep reading with a 7-day free trial
Subscribe to Core Memory to keep reading this post and get 7 days of free access to the full post archives.