Since 1999, two philosophers from Australia and the UK have proposed a thought experiment about the brain. In 2024, the hypothesis becomes a reality.
Based on the writings of philosopher and AI ethics expert, Ms. Dvija Mehta. She is currently affiliated with the University of Cambridge, working at the Leverhulme Centre for the Future of Intelligence.
Expert Dvija Mehta.
In March, a man named Noland Arbaugh demonstrated the ability to play chess using only his brain. Paralyzed for the past eight years, he regained lost capabilities through a brain implant designed by Neuralink, the company founded by Elon Musk.
“Controlling the cursor seems to have become my new instinct,” Arbaugh said during the livestream. “I stare at a point on the screen, and the cursor moves there.”
Arbaugh’s description suggests that he is the one directly using the mouse to control the chess pieces. But a question arises: is this action performed by him, or is it the chip implanted in his head that is executing it?
As a philosopher of human thought and an AI ethics expert, Ms. Dvija Mehta is intrigued by this question. A brain-computer interface (BCI) similar to Neuralink opens a new era where humans and computers become one; this new technology raises questions about identity, selfhood, and individual responsibility to the community.
In the near future, BCI technology offers a solution for disabled individuals like Arbaugh, but its applications could extend even further. Neuralink’s long-term goal is to enhance human capabilities for anyone using their products.
When a machine performs tasks that only the brain could do, can it be considered an action of the brain itself, or must it be viewed as an action of a separate entity from the human?
The Brain as an Extension
For decades, philosophers have pondered the boundaries of an individual: how far does the mind extend, and where is the line between the inner world and the external reality? At a basic level, one might argue that the mind resides solely within the brain, within a person’s body. However, some erudite philosophers have proposed more complex concepts.
In 1998, philosophers Andy Clarke (from the UK) and David Chalmers (from Australia) published a hypothesis about “the extended mind,” essentially claiming that technology can become a part of us. Using philosophical language, Chalmers and Clarke introduced a concept called “active externalism”, which suggests that humans can delegate their thought processes to an artificial object, integrating that artifact into their own minds through that action. The device acts on human command, allowing person and machine to function as a unified entity.
This proposal predates the advent of smartphones but captures the current context, where individuals delegate many cognitive tasks such as storage and calculation to devices.
Humans using machines as “extended accessories” – (Illustration).
If Arbaugh’s brain chip is not considered a part of his mind, a pressing question arises: does he truly control his actions?
Alongside active externalism, Chalmers and Clarke also presented a hypothetical scenario in which a person uses a brain implant to manipulate an object on a computer screen—similar to what Arbaugh demonstrated.
To play chess, Arbaugh imagines what he wants to do, whether it’s moving a knight or a pawn. His brain implant, the N1 chip designed by Neuralink, detects signals emitted from his brain and begins to interpret, process, and execute the action Arbaugh desires.
From a philosophical standpoint, what just occurred here? Is the brain implant a part of Arbaugh’s mind? If not, does Arbaugh control his behavior?
To understand why, consider the difference between two concepts: the event and the action. The event encompasses the entire mental processing of the human brain, including thoughts, beliefs, desires, imaginings, calculations, and intentions. The action is what the body executes, like clicking on this article.
Typically, there is no gap between the event and the action. Imagine an individual named Tuan playing chess. Without the aid of a brain-computer interface, Tuan forms the intention to move a chess piece in his mind and then directly executes that action with his hand. In this case, the event and the action are a seamless process; the action of moving the chess piece originates from the very event formed in the brain.
However, in Arbaugh’s case, he had to imagine his intention, and the brain chip executed the action affecting the outside world. Here, the event and the action are separate.
Illustration.
This separation raises many concerns, such as whether a person using a brain implant can fully control the actions generated from the chip. The brain and body can already be inconsistent in many actions, such as sneezing or pupil dilation—actions that occur spontaneously without brain signals—so how different could actions generated from the implant be?
Could the foreign chip become a parasitic entity, powered by batteries, eroding the voluntary will of the individual?
Philosopher Mehta refers to this as the “intentionality conundrum”, originally termed “contemplation conundrum.” In Arbaugh’s case, he overlooked a crucial component in the long chain of the thought-action process: the active hand motion executing the mouse action. What could happen if Arbaugh quickly changed his decision right before the chip was about to execute an action? And would the chip consider one of the hypothetical scenarios Arbaugh envisioned as an action to be executed?
In other words, factors such as the chip’s latency and the priority of a thought leave us uncertain about whose action is being taken.
On an online chessboard, the risk of mistakes due to the chip is relatively low, but when applying brain chips in high-stakes environments, who bears the blame for errors? If a brain chip causes a surgical error resulting in a patient’s death, who is responsible?
Illustration.
This is not the only humanitarian issue surrounding brain-chip interface technology. In the renowned science fiction novel Neuromancer by William Gibson, implanted devices can strip a person of their identity, even manipulating their body and unlawfully accessing something that no one should be able to touch: an individual’s thoughts.
The issue of “intentionality conundrum” will become more pronounced when a chip cannot clearly distinguish between “events that occur only in imagination” and “events in imagination that lead to action.”
Using the language of neuroscience, it is almost impossible to clearly delineate between imagination and intention. A 2012 study conducted by a team of neuroscientists showed that no definitive signal appears in the brain indicating where “intention leads to action.”
If it is impossible to determine which intentions will become actions, then a brain implant will also be unable to distinguish which actions should be taken. This makes the chip partially responsible for actions dictated by the brain implant.
Philosopher Mehta argues that as the hypothetical situation proposed by Chalmers and Clarke has become a reality, we should delve deeper into the ideas they have raised to connect the concepts of events and actions in a brain-implanted individual. Applying the hypothesis of the extended mind would ensure that individuals like Arbaugh take full responsibility for their actions, rather than sharing that responsibility with the chip.
Chalmers and Clarke argue that to experience as a single subject, an individual must think as a single subject. In other words, a person with a chip must consider the chip to be a part of their self, integrated within their inner world. This way, feelings of control, ownership, and responsibility can be fully realized.
Brain-Computer Interface Technology opens a new door, not only for individuals with impaired mobility but also for the field of philosophy to discuss the boundaries between mind and machine. As Chalmers and Clarke once remarked: “Only when the leadership of flesh and brain is usurped can we gain clearer insight into our true nature as beings belonging to this world.”
