November 2004

Picking Your Brain

Bioethicist Paul Wolpe explores the implications of wiring computers to the human brain.

By Erika Jonietz

Paul Root Wolpe

Position: Professor, Departments of Psychiatry, Medical Ethics, and Sociology, and senior fellow, Center for Bioethics, University of Pennsylvania; chief of bioethics, NASA

Issue: Brain-computer interfaces. Neuroscientists and engineers are developing technologies that allow the brain to interact directly with computers, from chips that could enable amputees to control prosthetic limbs to devices designed to enhance brain function. How will these new technologies influence daily life?

Personal Point of Impact: A founder of the field of neuroethics, which examines the implications of emerging neurotechnologies. Organized the first series of meetings on the topic in 1999 and 2000, bringing together leading brain scientists such as Steven Pinker, Steven Hyman, and Michael Gazzaniga.

Technology Review: A company called Cyberkinetics received U.S. Food and Drug Administration approval in April for a clinical trial of a brain implant designed to allow paralyzed patients to interact with a PC. Is the technology really advanced enough to make this sort of test ethical?

Paul Root Wolpe: There are issues with device testing of this kind in terms of human-research protection. The kinds of people that these devices tend to be tested on are deeply coerced by the nature of their disabilities. I don't think it's insurmountable; all medical progress depends on somebody being the first one to try a new technology. What is crucially important is really good oversight and really good informed consent. Given the history of oversight and of informed-consent issues with medical devices, it does concern me that these technologies will be used without strong external review and monitoring.

TR: Isn't this technology at a much earlier stage than where you would test a drug?

Wolpe: Generally in bioinstrumentation, that happens. For historical reasons, we are much, much more concerned about people ingesting drugs than we are about subjecting them to bioinstrumentation, and we have different regulations about how to test them and protect subjects. Pharmaceuticals alter the basic chemistry of our bodies; bioinstrumentation, until recently, was primarily external to our bodies. The problem is, the nature of bioinstrumentation is about to change, and emerging biotechnologies will be incorporated into our bodies much as pharmaceuticals are. Ten, 20, 30, or 50 years from now, perhaps, nanotechnology will develop little nanobots that are injected into our bodies to roto-rooter out our arteries. Would these be a drug or a bioinstrument? We need to begin to change the way in which we think about bioinstrumentation in general. We have to rethink our tendency to be less rigorous about applying bioinstrumentation to the human body than drugs. Right now, even these new technologies that may have profound effects on our brains do not have the degree of oversight that drugs do.

TR: But the payoff seems huge.

Wolpe: For people who are paralyzed, the Christopher Reeves of the world, the ability to manipulate things in the world with the mind is an extraordinarily desirable outcome. Implants for people who have locked-in syndrome -- so they can't communicate with the outside world -- are being tested right now and allow the subjects to directly translate brain impulses into computer responses, such that they can move a cursor around a screen and choose phrases, simply through thought. That is certainly a wonderful thing. It would be churlish to say, Let's not allow this person to communicate because we're not sure what the long-term effect is of putting electrodes in his brain. You have to ask yourself the risk-benefit question. But those cases are different than neurotechnologies that might eventually become fairly common.

TR: What kinds of technologies are those?

Wolpe: A lot of the technologies we're talking about are communication technologies; they take information from the brain and externalize it for one reason or another. We also have internalizing technologies -- cochlear implants, optic-nerve implants -- whose purpose it is to take information from the outside and give us access to it. These two technologies will eventually come together, and then we'll have interactive-chip technologies, such that we'll have input-output interactions.

But in terms of what most people mean by brain-computer interfaces, there's a lot of work being done to create noninvasive BCIs by putting electrodes on people's scalps or having them wear these caps that are infiltrated with sensors. The goal is a system that could retrieve much more detailed and specific information from the brain so that people could do sophisticated kinds of work through thought alone. It's very promising for people who are paralyzed, but it also means that I could sit here at my computer with a cap on my head and answer the phone, "type" on my computer, be connected to my colleague in the office next door -- through brain impulses alone. That's one area I think technology may take us over the next 50 or 60 years. We're going to be able to manipulate any system that has a sophisticated chip in it, everything from your wristwatch to your car.