- author, Dvija Mehta
- scroll, BBC Future
In March, a man named Noland Arbo demonstrated that he could play chess using only his mind.
“For me, it has become intuitive to imagine the cursor moving,” Arbault said in a video broadcast. “I just look at one place on the screen and it moves to where I want it to go.”
Arbo's description refers to a sense of self-agency. It is suggested that he was responsible for moving the chess piece. But who actually carried out these actions, the mother of the plant?
Brain-computer interface (BCI) technologies, such as those proposed by Neuralink, symbolize a new era in the interconnection between the human brain and machines. They invite us to reconsider our intuitions about identity, self, and personal responsibility.
In the short term, the technology promises many benefits for people like Arbo, but the applications could extend even further. The company's long-term goal is to make these implants available to the general population, which will enable them to increase and enhance their capabilities.
If a machine can perform actions previously reserved for the brain matter inside our skull, should it be considered an extension of the human mind or something separate?
Extended mind
For decades, philosophers have been debating the limits of personhood: Where does our mind end and the outside world begin?
At a primitive level, you can consider that our mind is located within our brain and body. But some philosophers have suggested that this definition is a bit more complicated.
In 1998, philosophers David Chalmers and Andy Clark presented the “extended mind” hypothesis. They suggested that technology could become part of us.
In philosophical language, the two scientists proposed active externalism – a way in which humans could delegate some aspects of their thought processes to external artifacts, which would be integrated into the human mind itself.
This proposal appeared before the advent of the smartphone, and worked to predict how our devices might assign cognitive tasks, from finding ways to get somewhere to our own memory.
As an intellectual exercise, Chalmers and Clark also envisioned a “sci-fi” scenario in which someone with a brain implant manipulated objects on a screen — actually something very similar to what Arbaugh did recently.
To play chess, Arbo imagines what he wants, such as moving a pawn or an elephant. The implant, in this case, the Neuralink N1, selects the neural patterns of your intention, before decoding, processing and executing actions.
Now that it has actually happened, what conclusion should we draw from this scenario, philosophically speaking? Is Arbo's implant part of his mind intertwined with his intentions?
If the answer is “no”, controversial questions arise to determine who is truly responsible for their actions.
To understand why, let's consider the conceptual distinction: events and actions.
Events bring together all our mental processes, such as our thoughts, beliefs, desires, imaginations, contemplations, and intentions. Actions are events that are being acted upon, such as your finger movements to scroll that report on the screen, at this exact moment.
Usually, there is no separation between event and action.
Let us take as an example a hypothetical woman, Nora, who plays chess. It does not have an integrated BCI.
By organizing events, Nora can form the intention to move one of her pawns forward one square, and she does so simply by moving her hand.
In Nora's case, intention and action are inseparable. She can assign the process of moving the pawn to herself.
But Arbo needs to imagine his intention and the implant carries out the action in the outside world. In this case, the event and the action are separate.
This raises important concerns. Can a person who uses a brain implant to augment his abilities maintain executive control over his actions integrated into a brain-control interface (BCI)?
Human brains and bodies already produce many involuntary actions, such as sneezing, coordination errors, and pupil dilation, but could actions controlled by implants appear to come from an external source?
Could an implant appear to be a parasitic intruder that would erode the purity of a person's will?
I call this problem the contemplation dilemma.
In Arbault's case, he eliminates crucial steps in the causal chain, such as the movement of his hand that accomplishes his chess move.
What happens if Arbo first thinks of moving his pawn one square forward, but in a split second he changes his mind and realizes that he should move it two squares instead of one? Or what if he was analyzing the possibilities in his imagination and the plant misinterpreted one of them as his intention?
On the chessboard, the stakes are low. But as these implants become more common, the issue of personal liability becomes more confusing.
What happens, for example, if an implant-controlled action causes injury to another person's body?
This is not the only ethical issue raised by these technologies. Its superficial marketing without fully solving the conundrum of meditation and other important questions could pave the way for a dystopia worthy of science fiction.
For example, William Gibson's novel Neuromancer highlighted how organ transplants can lead to loss of identity, manipulation, and loss of privacy of thought.
The key question in the contemplation puzzle is to determine when an “event of imagination” turns into a “deliberate imagination of representation.”
When I use my imagination to think about what words I should use in this sentence, it is an intentional process. Imagination directed toward action—writing words—is also intentional.
In terms of neuroscience, it is almost impossible to distinguish between imagination and intention.
A 2012 study by a group of neuroscientists concluded that there are no neural events that can be described as “action intentions.”
Without the ability to identify the neural patterns that determine this shift in someone like Arbaugh, we may become unsure of which imagined scenario is causing each effect in the physical world.
This would allow partial responsibility and authorship of the procedure to be attributed to the transplant. Hence, we wonder again whether these actions are really his actions, and whether they are part of his character.
But now that Chalmers and Clark's thought experiment and the extended mind have become a reality, I propose to reconsider their basic ideas as a way to remove the separation between events and actions in people with brain implants.
Adopting the extended mind hypothesis would allow someone like Arbo to take responsibility for his actions without sharing them with the implant. This epistemological view suggests that in order to have your own experience, you have to think of it as your own.
In other words, the person must think of the implant as part of their self-identity, within the confines of their inner life. This can lead to a sense of activity, ownership and responsibility.
There is no doubt that brain transplants, such as those performed by Arbault, have opened new doors for philosophical discussions about the boundaries between mind and machine.
Discussions about action and activity have traditionally centered around the boundaries of identity, between the skin and the skull. But with brain implants, these boundaries have become flexible, meaning that the self can extend ever further toward technology.
Or as Chalmers and Clark note: “When the dominance of skin and skull is usurped, we may be able to observe ourselves more honestly as creatures in the world.”
* Dvija Mehta is a philosopher of mind and AI ethics researcher at the Leverhulme Center for the Future of Intelligence at the University of Cambridge, UK.
“Web geek. Wannabe thinker. Reader. Freelance travel evangelist. Pop culture aficionado. Certified music scholar.”
