The recent announcement and the demonstration of implantable medical device from Neuralink has generated broader discussions and debate in the industry. Perhaps it is best to begin discussing about the neuroscience start-up, Neuralink, by placing the significance of its objective and the use of its brain-computer interface (BCI) within history and the evolution of various social formations. After all, the BCI is a product of scientific and societal evolution. It combines ‘methods, approaches and concepts derived from neurophysiology, computer science and engineering’ with the aim of creating interfaces that symbiotically links the internal cognitive and affectual components of the ‘living brain and the artificial actuators’ (Lebedev and Nicolelis, 2017). Neuralink emerges out of the intersections and developments of social life, economy and technology. For Elon Musk, the intricate composition of these factors has created new urgent, if not momentous, socio-economic and ontological necessities.
As such, Neuralink presents its BCI as a remedy to these micro, meso and macro uncertainties. Two presuppositions about the upcoming super-intelligent system of production and society has urged Neuralink to invest $158M dollars in the project:
- 1. The ‘existential or civilizational threat’ from super-intelligent machines or Artificial Intelligence.
- 2. The necessity to protect large swathes of the homo sapiens from becoming what Yuval Harari in his seminal book Homo Sapiens (2011) calls the ‘useless class’.
In the singularitarian and techno-nirvana view of Ray Kurtzweil and Peter Diamandis versus the possible civilizational extinction and scepticism of Nick Bostrom, Bill Joy and Jaron Lanier, Musk sides with the latter group. He recently declared: ‘If humans want to continue to add value to the economy, they must augment their capabilities through a merger of biological intelligence and machine intelligence. If we fail to do this, we’ll risk becoming “house cats” to artificial intelligence.’ The upcoming virtual super-intelligent system of production and Artificial Intelligence will require new needs, skills, merits, processes and tools that are fundamentally intellectual and psychological, as opposed to physical. And, for Musk, it is Neuralink’s BCI that will be this new scientific and innovative technology tool—is central to human survival.
In a way, BCI and the cyborgification of the modern human is akin to fire for the early homo sapiens. Just as fire led humans to survive unfriendly environments via the indirect development of language, cognition and skills, BCI will play the same role in the future transmutation of humanity. The only difference is that fire was never a universal existentialist threat which the machines can be.
For this reason, Neuralink’s ambitious goal is to ensure humans are able to read and write thoughts, memories, feelings and to telepathically communicate. The ultimate teleological stop for Neuralink is to merge humans with AI.
A set of urgent questions need to be addressed in order to really comprehend the significance of Neuralink’s BCI. What exactly is BCI and its trajectory? Is Neuralink different to previous BCIs and if so, how? What are the ethical and socio-economic risks of BCI?
What exactly is BCI and its trajectory?
BCI is one of the most rapid and expansive areas of scientific research and development. The theoretical work on BCI began in 1929 when Hans Berger contemplated and put forward the possibility of ‘“reading thoughts” with the electroencephalogram (EEG)’. He speculated that this was possible via ‘processing human EEG waveforms using sophisticated mathematical analysis’ (T Zander and J Bronstrup, R Lorenz, 2014, p: 66).
The first concrete phases of BCI research began at the University of California in the 1970s. It was Jacques J. Vidal (1973, 1977) who gave birth to the name Brain-Computer Interface (BCI). In the early days of the research and development, the focus was fundamentally on neuroprosthetic applications emphasising on restoring sight, hearing and movement. By mid-1990s the field began to see the first neuroprosthetic devices for humans. The ‘first wireless intracortial’ BCI was built in 1988 by Philip Kennedy and his colleagues ‘by implanting neurotrophic cone electrodes into monkey brains’ (Gohel, 2015, p: 8)
Zoë Corbyn in her article Are brain implants the future of thinking? writes that currently there has been an;
…injection of Silicon Valley chutzpah [which] has energised the field of brain-computer or brain-machine interfaces in recent years. Buoyed by brain gate and other demonstrations, big name entrepreneurs and companies and scrappy start-ups are on a quest to develop a new generation of commercial hardware that could ultimately help not only people with disabilities, but be used by all of us (2019).
There is then wide consensus that the BCI and the radical withering away of the archaic and natural divide between subjective inner self and objective reality is the future.
Why is the BCI important?
Most of the field experts and researchers recognise that the symbiosis of machine and human, organic life and technology is far more powerful than just the rigid, mechanical one-dimensional Artificial Intelligence. Neuroprosthetics allows the combination of sense-perception (feelings, emotions, etc.) and machine-enhanced rationality. The idea is that thoughts, feelings and various cognitive operations can be accounted for via the electrochemical signals that flow around and between the variegated network of neuronal circuits (The Conversation, 2020).
Each cognitive and psychological process is reduced down to what is called an ‘action potential’. These brain signals related to specific activities allow for the BCI to calculate, codify and organise neuronal information based on certain parts of the brain. This ‘biomimetic’ decoding allows not only for the mapping out of the electro-chemical activities but also a bidirectionality that permits for the enhancement and manipulation of these signals. The neural electrical impulses are translated into machine-readable information which in-turn can be manipulated for specific purposes. This way, BCI and neuroprosthetics allow for the restoring of vision and sense of hearing, detecting and preventing epilepsy, repairing stroke- induced injuries, restoring memory to Alzheimer patients, reinstating movement in paralysed limbs, and enhancing sensory and cognitive capacities.
To put it simply, BCI enables the production of a granular understanding of brain networks and activities. It also allows for the enhancement of the abilities of our brain through machine intervention.
There are numerous companies and enterprises which are currently working to build high-resolution brain-machine interfaces. Utah Array, Synchron, Paradromics, Kernel, Ctrl-labs, BrainGate and the social media giant Facebook are just a few to mention.
But it is Musk’s Neuralink that has really come to be a dominant force in the race to connect humans to computers via machine interfaces.
Is Neuralink different to previous BCIs?
Neuralink’s demonstration this month included a pig named Gertrude which was surgically implanted with a brain-monitoring device. Two things worthy of mention here: first that the experiment demonstrates that data was being extracted and collected via Neuralink’s device located in the brain of the pig. Secondly, that Gertrude was able to ‘move around while the implanted chip collect[ed] the data’. This encoding of brain activity is a grand step forward.
Yet, Neuralink is much more ambitious and radical in its ‘vision, software capacity and implantable hardware platforms’ (K. Wyggers, 2020). In the abovementioned demonstration, Musk announced that the company currently has ‘several ongoing studies with human participants’ and will go on to prepare for the first human trials in 2021. The proposed human trials have to adhere to FDA’S regulatory and ethical standards that are in place to provide the necessary checks and balances in the process, however, the process of transferring the neural implant from animals to humans are the same. Micron-scale threads are inserted into areas of the brain that control movement. Each thread contains many electrodes and connects them to an implant called the Link. The Neuralink device currently collects data from a limited area of the cortex and covers a limited number of neurons. Important brain functions utilize many parts of our brain simultaneously such that the device has to expand to cover other areas of the brain to reach its full potential.
The company’s goal is three-fold: 1) treat brain disorders and aid people in need of urgent medical attention, 2) create a BCI and, 3) manufacture a ‘sort of symbiosis with artificial intelligence’.
Neuralink has been able to make leeway in key issues and factors surrounding BCI. These include:
- 1. The complete automation of the insertion of the in-brain device via an eight-foot surgical robot. The Neuralink demonstration did not cover how the robot-inserted microelectrodes get into the cortex and find cells. This is the most critical and difficult part of a BCI product.
- 2. Building an in-brain device (version 0.9) that is surgically placed in the skull.
- 3. Conventional implants are stiffer than actual brain tissue which cause error and instability in recording or damage to brain tissue. Neuralink has produced ‘flexible BMI’ which are ‘much softer than silicon implants and similar to actual brain tissue’.
- 4. The existing BMI systems pass through skin and require laboratory equipment and personnel. Neuralink’s device relies on safe and effective home operated clinical system.
- 5. Neuralink’s capacity to listen to brain cells and enhance its signals is radically higher than its closest competitor. Whilst BrainGate implants arrays of 100 electrodes in the brain of paralysed patients, Neuralink has built a device to relay signals wirelessly from 1024 electrodes.
- 6. Neuralink’s procedures have minimised damage to brain tissue during implantation and operation.
- 7. Neural spikes contain a lot of information. Decoding and interpreting this information has thus far been a difficult and inefficient task. Neuralink labs have designed computer ‘algorithms controlling a virtual computer mouse from the activity of hundreds of neurons’. The device is able to connect and control much more neurons
What are the ethical and socio-economic risks of BCI?
The ethical implications of the BCI are too great and extensive for the scope of this article. However, three key issues need to be briefly touched upon: the unequal access and utilisation of the device, cyber-attacks and privacy. The first involves the geographical and economic inequality in the production, distribution, consumption and operation of the BCI. The unequal access to this device in an already highly fragmented and polarised world could lead not only to mass socio-economic divides (between the useless class and algorithm/device owners) but also to the literal social and biological split in the species. If this process is not democratic and regulated, we may see an absolute division between a super-intelligent cognitively augmented cyborg, and the limited and primitive homo sapiens.
There is also the risk of security in storing and transmitting information. The bidirectionality of the signals and the linkage of the living brain to machine leads us to a rational equation: if the signals can go out and be enhanced they can also easily come in and be integrated. The signals and the algorithms to decode the signals can easily be hacked into by foreign entities. Here, Pablo Usieto and Javier Minguez ask a set of questions that need to be analysed and, in the long term, answered:
What would happen if a computer hacker could access the neural information of the president of the United States and revealed the nation’s next military objectives? Or what if the hacker discovered the president suffers from a mental illness and disclosed this information to the press? Or even worse… How much would a rival country pay for such information? (Usieto and Minguez, 2018)
The question of the control of the technology and regulation of its use becomes paramount here.
Perhaps, the last key issue that needs a mention, and linked to the previous point, is the matter of privacy and drawing the lines of access and control which the BCI has over our brains. How far should the BCI access our thoughts, ideas and emotions? How can certain ideas and thoughts be protected from the reach of the device? Which ideas are to be protected and which must be open to the social world, if any? Currently, the subjective and objective areas are demarcated and tightly defined, yet with the emergence of the BCI, bionic and biological innovation, this divide is radically pronounced. This will have dire effects on subjective autonomy and sovereignty of our lives. It is why Musk is fearful that we may be ‘summoning the demons’. We are talking about humans losing control over which ideas can be thought and which ideas can be shared with self-replicating and autonomous computers, society, market forces and the state.
In the evening of modernity, René Descartes, the French philosopher and mathematician, famously wrote cogito ergo sum (‘I think, therefore I am’). This maxim summarises the historical conception of subjecthood since the evolution of the species from the greater apes. It also encompasses the relationship we have had as species with ourselves, with others and our environment. There was always the ‘I’ then there was the other and objects (the natural). This is all about to change. A successful autonomous and bidirectional BCI system will alter the trajectory of human development. The application of the technology transforms us, and our world, beyond our imagination and comprehension. Our thoughts and beliefs will be made public and transferrable and the outside world will constantly intrude the inner kernel of subjective autonomy, the soul.
With the emergence of the BCI, the ‘I’ in Descartes’ maxim will be very unclear in the coming future.