Image courtesy of Needpix.com

“I’ve begun to wonder what’s me and what’s the depression, and what’s the stimulator…it blurs to the point where I’m not sure…frankly, who I am.”

Imagine the following cases: A person with paralysis is able to control a robotic arm, using his thoughts alone, and it allows him to reach and get his coffee cup. A person with Locked-In Syndrome is able to communicate with her family for the first time in five years. A teenager struggling with anorexia regains weight after her body dysmorphia dissipates. A war veteran who lost a limb is able to use a new haptic-enhanced robotic arm to feel her child’s hand in hers again. These examples illustrate the breadth of possibilities that brain computer interface (BCI) technologies offer our society. A BCI is a device that creates a communication channel between the central nervous system and a processing device. This device collects neural data, analyzes it using algorithms that can be artificially intelligent, and creates an output that is sent to another device, like a robotic arm, a computer cursor, or a stimulating device in an area of the brain, such as a deep brain stimulator (DBS).

What makes BCI different from other neural implants, like DBS, is that they “read” neural activity and use this information to modulate output, compared to a fixed stimulation input. This extra layer, however, raises concerns about the ability of BCI technologies to influence (e.g., enhance, diminish, or confuse) human agency. Recently, Elon Musk announced that his company Neuralink intended to implant a human with a BCI device involving thousands of individual electrodes by the end of 2020. We must ask whether BCIs should be introduced into society broadly given their capacity to alter people’s sense of agency, or control over their actions and thoughts. 

Image courtesy of Pixabay

Human agency is the ability to act and take ownership of experiences and actions. Agency gives people a sense of control over their minds and bodies and is an important part of the human experience. But agency is a complicated phenomenon that relies on or is constituted by at least several other features, including responsibility, privacy, authenticity and trust.  People want to feel that they are intentionally acting rather than merely being acted upon. This means that when a person uses a robotic arm to reach their coffee cup, they feel responsibility over the actions of the device. It means that when a person with Locked-In Syndrome types out of a message using a BCI, she does so from a background of privacy, such that her thoughts and body are only accessible to those with whom she shares them.  It means that the teenager with anorexia feels that the BCI does not undermine her authenticity, that she continues as a recognizable self. Finally, it means that the war veteran can trust that the feel of her child’s hand is reliable and not misleading.  In our view, these concepts of responsibility, privacy, authenticity, and trust are crucial to understanding human agency and are uniquely implicated by different types of BCI technologies.  

As the examples above show, BCI technologies have the ability to both improve and diminish human agency. While a BCI may allow a person with anorexia to change the way she perceives her own body, it may also lead to changes to her personality that are unwanted or create anxiety. Similarly, when a paralyzed person uses a robotic arm to shake someone’s hand, if the device instead crushes that person’s hand, the user may be unsure of his responsibility for the action. Because different BCIs may implicate different aspects of agency in different types of end users, understanding and categorizing the problems that can occur is key. Failure to understand these issues could inhibit meaningful informed consent, end user acceptance of BCIs, measurement of agency, and user-centered design.  

We have been awarded a grant from the NIH to study these issues of agency. In our R01 grant, we aim to develop a conceptual map of human agency, using both philosophical literature on the relevant features, and end-user input on the phenomenal experience of using BCIs in research studies, through longitudinal interviews with participants using motor, sensory, communication, and psychiatric BCIs.  

Image courtesy of Pixabay

Unlike previous studies, we are exploring BCI technologies across modalities. Many of the early studies around end user perspectives for DBS were isolated to one disease condition. Understanding experiences across our four modalities—motor, sensory, communication, and psychiatry—will allow us to consider, for instance, whether privacy matters in the same ways to users addressing issues of communication compared to those treating motor impairments, or whether concerns about authenticity arise in the context of movement in similar ways to how they come up in the psychiatric realm.  

If we are successful, we believe this research will go a long way in guiding the field of neuroethics in developing better mechanisms to protect end users from potential unwanted consequences of BCI devices, while encouraging the design of BCIs that promote user agency. This includes protocols that warn future users about the potential changes to agency they might experience and to help them weigh the negative risks of a device against the potential therapeutic benefit. We also believe that working closely with BCI scientists will help develop a research enterprise that is better able to identify ethical issues raised by future technologies and utilize end user perspectives to design the next generation of BCIs.

Recommended readings 

______________