BCI Samara 2020 Neuralink Panel Discussion Overview

By Maria Marano and Yannick Roy

Image Courtesy of bcisamara.com
This article is based on a recent panel discussion entitled “Neural Implants - Neuralink… but not only,” featuring Jonathan Wolpaw (NCAN), Karen Rommelfanger (Emory University), Mikhail Lebedev (HSE), Xing Chen (Netherlands Institute for Neuroscience), moderated by Yannick Roy (NeuroTechX) that took place as part of the BCI Samara 2020 exhibition. The full panel recording can be viewed here (begins at 6:31:51).

On October 9th, some of the foremost thinkers in the field of neural implants and neural ethics convened to discuss this summer’s Neuralink update. The conversation spanned across numerous topics, reflecting the depth of expertise on the panel. Throughout the discussion a consistent theme emerged: the importance of present-day planning aligning with future goals in the development of neural implants. At times, the panelists both commended and expressed concerns about the alignment of Neuralink’s plans with its long-term vision. Below are some of the key points discussed:

Tech Specs

Researchers almost universally agree that the science presented at the Neuralink event was generally conventional and on-par with the current field of neural implants. Conversely, the technological advances presented were impressive and applauded by the neurotechnology community. The Neuralink prototype (known as “the Link”) is a small, thin, fully implanted device with no cumbersome external wires and an inductively rechargeable battery. While standard models like the Utah array typically incorporate a channel count in the hundreds, the Link boasts an impressive 1000+ channel count within its small frame. Beyond the device itself, the Neuralink team is developing a robotic surgery platform. The “sewing machine” robot is based on a model developed at UCSF and would allow for the safe and precise insertion of the Link’s fine electrode threads while avoiding blood vessels. 

Image Courtesy of Flickr
Despite these impressive technological gains, many of the panelists conveyed caution in their excitement for the tech. While the higher channel count is a significant increase from typical models, Dr. Lebedev discussed how a high-channel-count would only provide a temporary benefit as the quality of recordings diminishes over time. Currently, biocompatibility is an unresolved challenge in the field of neural implants. As electrodes constitute foreign objects in the brain, an immune response is mounted following implantation that damages the tech. Dr. Chen acknowledged that the Neuralink team is well-equipped with experts to address this specific problem. Nonetheless, at this time, it is unclear how the Link would avoid these complications, and until this issue is resolved, the high channel-count may be of limited value. Ultimately, while a higher channel count may improve resolution, Dr. Chen offered a more pragmatic target -- using the minimum number of channels necessary to improve patient quality of life. 

Beyond biocompatibility, it is still unclear if additional channels are the best approach for an optimized device. Some experts in the field propose that other recording measures (i.e., local field potentials through ECoG) may provide more stable values. A neuron-only focus may also be too limited, depending on the long-term aims. Overall, the panelists stressed the necessity for a roadmap of concrete goals and targets in the ongoing tech development

Animal Models

Let’s be clear: the star of the Neuralink presentation was Gertrude, the pig. During the event, Elon Musk demonstrated the Link’s abilities both live and through pre-recorded video using pig test subjects. Prior to the event, scientists speculated if the demo would involve human participants, and many were disappointed to discover the science had not advanced as far as Musk’s hype perhaps warranted. 

An important consideration in research is the appropriateness of the model subject for the experiment. The principles of the ethical use of animals focus on the three R’s: reduce, replace, and refine. In discussing the choice of a pig model, the panelists again emphasized that selection should be based on the eventual goal of the project. Each animal model carries different strengths and weaknesses. Rats, for example, are effective in a stimulus-behavior output model as they learn reward systems quickly and often respond more reliably than primates. Likely, pigs were selected by the Neuralink team for their high intelligence. Compared to pigs, primates tend to be more costly, have more ethical uncertainty associated, and are generally more difficult to manage in a research setting. 

Another important consideration for the ethical use of animals in research that pertains to invasive neural implants is the animal’s suffering. It is important to select the animal model with the lowest pain perception capacity and include the fewest number of animals possible to meet the intended research goal. Animals must be safely anesthetized during surgery, but researchers must also look out for any signs of pain or suffering following the procedure, as pointed out by Dr. Rommelfanger. A big concern is the risk that animal suffering may go undetected by scientists, and important new methodologies and techniques are being developed to address this concern. 

Elon Musk: The Personality behind the Science

While the authors of this article maintain that Gertrude was the star of the Neuralink update, we concede that Elon Musk was the main draw for the event. Idolized by as many people as he is despised (and an internet meme to the rest), Musk has solidified himself as an iconic figure in modern-day society. While his notoriety has benefits, as it directs public attention and funding to important areas of STEM (electric vehicles, space, and, more recently, neurotechnology), it also runs the risk of eclipsing the work of similar companies and diverting interest and recognition from the scientists who built up the field before him. While many other neurotechnology companies have equally innovative and exciting products in development, their public launches are unlikely to garner 2 million views. As part of industry, Musk is not encumbered by the roadblocks of academia, such as ethics approval delays or the necessity for publication. While this can greatly expedite progress, further contributing to the aura of exceptionalism around him, it also means important safeguards, like the peer review process, go unchecked. 

Musk has captured his audience with his boldness: both in the sky-high goals he sets for his organizations and his ability to enter, and seemingly master, completely new fields of STEM. But it is important to reflect on the potential ramifications of the Elon Musk persona. In addition to overhyping his Neuralink update, Musk also tends to set overly ambitious deadlines for himself and his teams. Just before the August update, articles were released describing a rushed work environment. Musk also garnered attention for setting grandiose expectations on the popular Joe Rogan podcast earlier this year, where he claimed (among many other things) that in “maybe 5-10 years, best-case scenario,” there would be a “universal language through computers.” Given that Musk’s own neural implant has only debuted in animal models thus far, this is a daring claim of neurohype. Nonetheless, he regularly speaks with this level of imprecision about both the near-future availability of his devices and the numerous illnesses it will be able to treat. While aggrandized claims about luxury cars may not be a huge issue, hyperbolic speech in relation to healthcare can seriously impact patient communities. Buried within almost every tweet from the Neuralink account are requests seeking cures to neurological conditions or individuals offering themselves as test subjects. Given his platform, there is a cause for concern about the ethics around his exaggerated marketing. Is it morally acceptable to publicly promote potentially life-changing therapeutic interventions without concrete evidence supporting those claims?

Image Courtesy of Twitter

AI Threat & Digitalization

In addition to curing neurological conditions, another major impetus for the founding of Neuralink was Elon Musk’s concerns about AI. Specifically, he believes that AI will surpass human intelligence and could be used (or go rogue) against humanity. This is a concern shared by other thought leaders, and Musk is one of the many signatories on an AI Principles document through the Future of Life Institute. Musk argues that connecting our brains to computers through neural implants will allow humans to harness the same computing power as AI and stand a fighting chance against a threat he sees as “far more dangerous than nukes.” 

The lively discussion around the urgency of the “existential threat of AI” was one of the most thought-provoking sections of the panel discussion. Dr. Rommelfanger discussed the “cult of personality” around Musk and his tactic of playing into fears as a way to promote his projects (i.e., need to colonize new planets, the threat of AI, etc.). Dr. Rommelfanger continued by explaining how existential fear around technology has been a consistent theme across all major technological advances in human history and how thought-leaders will often use these fears to garner authority. Dr. Wolpaw countered that previous advances had not involved the nervous system, and concern is much more warranted in this case. The two reflected on whether we should focus on planning for long-term threats, like superintelligent AI, or prioritize existing/more short-term threats, such as privacy concerns around neurotechnology and unintended consequences of neuromodulation

While Neuralink insists its primary focus is solving neurological illnesses, comments about universal computer languages and combatting superintelligent AI suggest much broader public uses. Musk has mentioned that Neuralink will be available to the general public—“but not a mandatory thing”—without addressing issues like disparities in accessibility to neural implants and rising fears from targeted individuals in the growing neurotechnology space. As it stands, our limited understanding of human neurobiology means it will likely be some time before Musk and his device are capable of producing therapeutic benefits or self-enhancement. Nonetheless, it is an incredibly exciting time for the field, so much so that when the panelists were asked for their sci-fi recommendations, the group struggled to offer suggestions. Dr. Wolpaw summarized it best in saying, “real science is much more exciting than science fiction.”

Additional Resources

___________

Maria Marano works as an health educator at a non-profit and is co-lead of the NeuroTechX Diversity Initiative. Maria’s focus is in the field of neurodegenerative diseases and working to bridge the gap between research discoveries and public understanding.


Yannick Roy is co-founder and executive director for NeuroTechX – the international neurotech community. Yannick is on a mission to accelerate the advancement of neurotechnology while shaping a responsible, inclusive, and accessible neurotech future. 

Want to cite this post?

Marano, M. and Roy, Y. (2020). “BCI Samara 2020 Neuralink Panel Discussion Overview”. The Neuroethics Blog. Retrieved on , from http://www.theneuroethicsblog.com/2020/11/bci-samara-2020-neuralink-panel.html

Comments

Follow Us

Follow Us
Emory Neuroethics on Facebook

Emory Neuroethics on Twitter

AJOB Neuroscience on Facebook