Neuralink: the company pushing the boundaries of AI implantation
Neuralink is a relatively recent venture into the neurotechnology field founded in 2016, by self-appointed tech magnet Elon Musk. Its main goal is to develop an implantable brain-machine interface (BMI) at a rate far greater than any other traditional industry leader. While much of its work remains carefully hidden, from the presentations it has put on and through interviews and papers from Musk and his team, some insight has been gained into what exactly they are planning to do.
Currently the most advanced FDA approved BMI is a parkinson's deep brain stimulator that only has 10 electrodes and a relatively low success rate, however there have also been other research projects demonstrating the use of the technology that neuralink plans to use. One of these was the Utah Array, which utilises 100 electrodes and has allowed people to connect to and use mobile devices and their functions such as texting.
Neuralink has one key project at the moment, which is the base for all their future endeavours, and which encompasses several different elements. As they are revolutionising the bio-implantation industry they are designing or redesigning all their components from scratch albeit learning from the previous work of others. The first major change are the threads that are being used. Historically they have been made from rigid metals or semiconductors which while relatively effective, have several disadvantages. Primarily they disrupt and irritate the tissue surrounding them which causes immune responses, leading to their function becoming limited and the populations of neurons around them restricted. Neuralink has therefore fully developed new, miniaturized flexible multi-electrode polymer probes. These are able to each take in dozens of different neurons’ electrical spikes created by action potentials and then record and analyse them. The average thread thickness is between 4-6µm and therefore creates very little inflammation in the areas in which they are placed into.
Along with this development into thread technology comes the method of implantation. Neuralink has developed an automated robot to carefully insert each thread individually with pin-point accuracy, so as not to damage blood vessels in the brain and lead to inflammation. It also allows for a greater concentration of electrodes in a smaller area. The robot uses a 24µm tungsten-rhenium needle allowing it to both grip the threads with strength and penetrate through the meninges and brain tissue. Throughout the entire operation though, a surgeon maintains full control and is able to make last minute adjustments to the insertion sites of threads. This ensures total avoidance of the vasculature.
The final but no less important part that is being developed are the electronics that link everything together. They comprise of several different hubs called N1 chips for all the threads and the individual electrodes coming off of them, and then a wireless connector which all these chips link to appropriately named ‘the link’. The N1 chip is 8mm in diameter and only 4mm tall, which is important as it gets implanted under the scalp as well. It then reads the signals from the threads, amplifies and digitizes them, processes them and finally sends them to the link. They are designed to be completely wireless, last decades and be implanted through an entirely non-intrusive surgery. The link will connect to an app on your phone for incredible ease of use, and any updates will only be to the link and app, no surgery will be required.
By the end of 2020 Neuralink hopes to have it’s first chips implanted into humans, having performed extensive animal tests, with their first goals being total control of your phone without touching it. As Musk has said many times, we have the world at our fingertips, but the bandwidth is too slow for it to be of any use - fingers and speech are rubbish input devices - and therefore the underlying goal Neuralink has in mind is to add a third layer to your brain through complete interconnectivity with computers and eventually AI.
Ricardo Pilnik 6F