Co-authored by Chris Chiang

In August, over 150,000 viewers watched a live demonstration showcasing the latest advancements from Neuralink, a company founded by Elon Musk focused on brain-computer interfaces. Their innovative approach involves a small brain implant aimed at merging human cognition with artificial intelligence.

The demo presented an intriguing mix of neurosurgery robots, real-time neuron activity from a living brain, and a memorable moment where Musk engaged a Tamworth pig named Gertrude, inviting her onto the stage.

Musk is known for his showmanship, often blurring the lines between tangible technology and futuristic aspirations. Neuralink itself is a multifaceted enterprise that intersects various fields, including medical technology, robotics, neuroscience, and AI, adding to the complexity of grasping its innovations.

With a background in cognitive science from Johns Hopkins University and years of experience in AI, I, alongside my colleague Chris Chiang, a biomedical engineering graduate from the same institution, began examining what Neuralink's demonstration truly indicated. What medical applications are viable? What developments have occurred since their last update in 2019? And what are the broader implications for brain-computer integration?

At the heart of Neuralink's technology is the Link, a compact implant that Musk likens to a "Fitbit for your brain." The Link charges wirelessly through electromagnetic induction and is designed to be inserted via a specialized surgical robot that operates similarly to a sewing machine. During the procedure, a small hole is drilled in the skull for the Link, which is then concealed beneath the scalp, making it undetectable.

The Link features over 1,000 tiny electrodes arranged into threads, each measuring approximately 5 microns in width—significantly thinner than a human hair. The surgical robot precisely places these threads into the brain's cortex, enabling the device to both read and send signals to the brain. Musk describes the insertion process as relatively non-invasive, with the robot carefully navigating around blood vessels to minimize any bleeding.

While the concept of a robot performing brain surgery may seem unsettling, the precision required for inserting the Link’s tiny electrodes necessitates robotic assistance. The device's design aims to avoid damaging blood vessels, which is crucial for maintaining the integrity of the brain's signaling capabilities.

Neuralink asserts that the implantation procedure would last about an hour and can be done on an outpatient basis without general anesthesia. Furthermore, the company claims the process is reversible, albeit leaving a small opening in the skull. Musk envisions users upgrading their Links periodically, akin to upgrading smartphones.

This raises both exciting prospects and significant concerns. The Link stands out compared to existing neural implants like the Utah Array, which requires external hardware and is linked to a bulky device on the skull. The Utah Array also carries substantial risks during surgery, including a 1% chance of stroke and up to a 5% risk of severe post-operative infections. In contrast, the Link's design minimizes exposure to pathogens by being entirely embedded beneath the skin.

The electrodes within existing implants are often rigid and restricted to specific brain areas, while the Link’s threads can distribute electrodes across multiple regions, potentially providing greater access to the brain's functions. With over 1,000 electrodes compared to the Utah Array's 256, the Link may offer higher data resolution, enhancing our understanding of brain activity.

Despite these advancements, Neuralink's plans for the Link have evolved since 2019. Initially, the vision included accessing deeper brain regions; however, the current focus is on interfacing with the cortical surface. This limitation means that many ambitious treatments Musk discussed are likely still years away, particularly regarding conditions like memory loss and mental health issues.

Nevertheless, significant potential exists within the cortical regions. For instance, the primary somatosensory cortex, which processes physical sensations, could be accessed to create a brain-connected artificial limb for amputees. By wirelessly linking sensors on a prosthetic limb to the primary somatosensory cortex, users could gain realistic tactile feedback.

Neuralink's demonstration included a pig walking on a treadmill, with brain signals accurately predicting leg movements. This suggests the device could greatly enhance fine motor control and sensation for users, surpassing existing technologies.

Moreover, Neuralink has hinted at using the technology to treat paralysis. The potential to restore movement and sensation to individuals with spinal injuries could be the first application of the Link in humans, a possibility that has drawn the attention of the FDA, which has granted the device a “Breakthrough Device” designation.

The Link could also connect with other brain regions like the primary auditory and visual cortices, enabling basic stimulation, such as creating sounds or visual flashes. This could serve practical purposes, such as alerting users to notifications.

However, to achieve more complex functions, like memory alteration or mood regulation, deeper understanding and access to various brain regions would be necessary. Neuralink aims to utilize machine learning to correlate brain activity patterns with real-world events, potentially enabling the device to interpret thoughts or even facilitate communication between users.

Despite these ambitions, significant challenges lie ahead. Existing research in machine learning and neuroscience has not yet fully defined how to interpret complex mental states or conditions like depression. Moreover, the intricacies of brain function present substantial hurdles for effective treatment.

Concerns about the long-term viability of implanted devices also persist. The Link's longevity will hinge on factors such as its hermetic sealing against moisture and the body’s reaction to a foreign object. Increasing the device’s size slightly could enhance its durability, while the materials used in its construction will also play a crucial role.

Beyond medical implications, the security and privacy risks of connecting a computer to the brain are profound. Potential threats include unauthorized access to thoughts or the possibility of manipulating brain functions, which could have dire consequences, especially in critical situations like driving.

While current implantable devices face security challenges, no major incidents of malicious use have been reported. However, the potential for exploitation in the future raises significant ethical questions about consent, advertising manipulation, and personal agency.

Neuralink will need to implement robust security measures to protect users, drawing parallels to existing cybersecurity practices. Addressing the FDA's requirements will be crucial for moving forward with human trials.

Musk’s presentation serves not just to showcase technology but to attract talent and investment for continued development. Although the company has yet to release a market-ready product, the potential applications of the Link are vast, particularly in treating serious conditions and enhancing human capabilities.

As interest in neural implants grows, the prospect of Neuralink's technology providing tangible benefits for users is increasingly plausible. While fully realized telepathy remains distant, advancements like the Link may soon offer significant improvements in quality of life for individuals with disabilities, potentially transforming our interaction with technology.