Brain-Machine Interface

From SI410
Jump to: navigation, search

Brain-Machine Interface is a type of technology that establishes communication between the biological brain with an external device, allowing the user to interface with said devices via neurological input. As this technology makes advancements ranging from new generation of prosthetics to neural interfaces for phones and browsers, large tech corporations are now developing their own BMI technologies. Considering the intimate nature of the data accessed by BMI, the gap between regulation and technological innovation, along with the track record of tech corporations' mistreatment of user data, it raises an array of ethical issues and concerns that have yet to be addressed nor discussed.

Neuralink's proposed module that sits outside the head and wirelessly receives information from threads embedded in the brain


Background

How it works

The reason BMI works in the first place is that the human brain is fundamentally compatible with machines due to electricity. The human brain consists of billions of neurons that are connected to one another via axons and dendrites. When the brain feels, senses thinks or acts, these neurons produce an electrical signal (caused by the controlled imbalances in molecular charges around the neuron) that passes through the large networks of neurons and eventually controls and communicates every behavior of the human body. BMI harnesses and relays the electrical signals produced by the neurons to an external machine which then interprets the signal to carry out a specific function[1].

Current Applications

Hugh Herr, the director of MIT biomechatronics group on stage with a dancer who lost her leg during the Boston Marathon bombing but regained the ability to dance professionally with his prosthetics

BMI has been in medical applications since the 1960s and was mainly used to rehabilitate biological functions that were lost or damaged. Cochlear and retinal implants are well known and established examples of BMI where the implant stimulates the damaged or degenerated sensory organs with electrical signals to simulate sounds or color. One of the more innovative applications of BMI technology today is in neural prosthetics, which has revolutionized the field of prosthetics. Traditionally, patients who need prosthetics were able to gain very little range of movement and dexterity due to standard prosthetics being rigid and unsophisticated in terms of possible movements. With neural prosthetics, it is no longer just a hunk of material, instead, it consists of hundreds of motors, actuators and sensors that allow wide range of motion in places such as the fingers as well as knees and ankles. Along with this biomechanics support, neural prosthesis receives and interprets electrical motor signals from the brain, and carries out specific movements that would have happened with the original biological limbs. Due to this, it usually takes minutes for someone to get used to this brain to prosthetics communication before they can start moving their artificial limbs naturally. [2]

BMI and Big Tech

Neuralink

One major player of this endeavor is Elon Musk, the CEO of Tesla and SpaceX among many other innovative ventures. His latest venture Neuralink has announced in 2016 that they are working on building a brain-machine interface called Neuralink, which will bootstrap and augment the capabilities of our biological brain by taking advantage of the advancements made in AI. The motivation for this is to compete with artificial intelligence, and to ensure humanity is equipped with enough brain(-machine) power to properly take advantage of AI, not get taken advantage by it [3]. In Musk's own words, humanity has an information bandwidth problem. While computers are able to process and communicate information at ever increasing rates, the human brain is limited by our biological infrastructure. In the time it takes our eyes to take in visual information and process it, a computer can process and parse magnitudes greater amounts of data. Not only that, it takes humans even longer to communicate what we processed, whether that be through speech or typing on a keyboard with our fingers. In the coming years, this gap's inability to process and communicate information will only widen, and it is not hard to see a society where humans are obsolete in many circumstances. Neuralink's goal is to harness the capabilities of AI and directly augment our own abilities with it to establish a symbiosis[4]. Musk states that pending FDA approval, Neuralink plans to have the first patient trial before the end of 2020[5].
Neuralink's app that teaches users how to control their phone with their thoughts

Facebook

Another tech giant Facebook has taken steps to establish themselves in the BMI sector. They aim to create a wearable device that will enable users to type and navigate the web through thought. The team working on this claims this will allow people to type five times faster than the speed at which we type on the phone, and it would also allow people who are paralyzed to communicate with ease [6][7]. Facebook has also acquired a brain-machine interface startup CTRL-labs for approximately a billion dollars to further bootstrap their BMI initiative. Considering Facebook already owns OculusVR, one of the largest platforms for virtual reality technology, their ability and position to integrate BMI technology into entertainment, as well as practical purposes is significant.

Ethical Concerns

Relationship between computers and its users

Over the years of innovation in computing machines, one can observe a pattern regarding the relationship between humans and computers. Starting with the Difference Machine invented by Charles Babbage in 1822, this "computer" took up an entire room, and was very much a separate entity from its users and hardly contained any personal data. Fast-forwarding to the next big leap in computer design was the introduction of the original Macintosh by Apple in the 80's. This was revolutionary in that it was a personal computer that could sit on the users' personal desk in their room, whereas before computers were mainly accessible in a lab or a university library. As close and convenient it was for the user, the Macintosh began to hold more personal and private information of its user. The pattern here is that as the computing device gets closer and closer to its users in proximity, more sensitive the data it gathers becomes. This trend continues with laptops, to smartphones, which began gathering locational data, and finally, wearable technology, which can gather biometric data such as heart rates and facial/retinal recognition. BMI is the next step in this trend, and thus generates data that is on a new level of privacy and intimacy.

Data Ethics

Brain-machine interfaces by nature require close proximity to the user because it needs to establish a link between the brain and the device. In the case of BMI that requires surgical procedures such as Neuralink, implants are placed directly on to the brain to establish a connection between it and a chip[8]. Therefore as private corporations begin to build out their BMI initiatives, they will have access to data that provides insights into the user's very thoughts, among other highly sensitive biological information.

This is of concern considering the track record of large tech corporations when it comes to their treatment and handling of user data, such as the Cambridge Analytica incident with Facebook[9], or the Apple court case regarding revealing user data to advertisers[10]. The intimate and biological nature of user data afforded by this new generation of BMI, along with incompetent practices of corporations raises the stakes of data harvesting and data handling practices. Even in the hypothetical scenario that these companies handle this data perfectly, they are still prone to attacks and breaches from malicious parities as was seen in the 2017 Equifax breach that exposed information of 147 million people [11]. What makes this more concerning is that the regulatory climate has not caught up with advancements in the internet and communications sector, yet the tech industry is taking another large leap with advancements in BMI. At the current pace, our regulatory system may not be equipped to properly assess the implications of these devices and the data they will afford.

References

  1. https://towardsdatascience.com/a-beginners-guide-to-brain-computer-interface-and-convolutional-neural-networks-9f35bd4af948
  2. https://www.youtube.com/watch?v=PLk8Pm_XBJE
  3. https://towardsdatascience.com/merging-with-ai-how-to-make-a-brain-computer-interface-to-communicate-with-google-using-keras-and-f9414c540a92
  4. https://www.youtube.com/watch?v=Vf6ZOiTxTi4
  5. https://www.cnbc.com/2019/07/17/elon-musk-brain-machine-startup-neuralink-plans-human-trials-in-2020.html
  6. https://www.wired.com/story/zuckerberg-wants-facebook-to-build-mind-reading-machine/
  7. https://techcrunch.com/2017/04/19/facebook-brain-interface/
  8. https://www.youtube.com/watch?v=mty_WVP8DvA
  9. https://www.nytimes.com/2018/03/19/technology/facebook-cambridge-analytica-explained.html
  10. https://www.theverge.com/2019/5/28/18643146/apple-itunes-privacy-listening-data-disclosure-lawsuit-rhode-island-michigan
  11. https://www.ftc.gov/enforcement/cases-proceedings/refunds/equifax-data-breach-settlement