Today, Elon Musk gave an update on Neuralink’s progress on building an implantable device comprising 1,000 electrodes that can send and receive electrical signals via bluetooth. The device can already be implanted by a robot, and has already been tested on pigs. The demo featured a pig with the implant and a visualization of the signals (from electrodes near the part of the brain that is stimulated by olfactory nerves) being received in real time as the pig was walking around, sniffing and eating.
Furthermore, the team was able to build a neural network that could predict the pig’s movement of its legs just from the signals being received from the brain. It is truly fascinating that neural nets could be used to map electrical signals in the brain to real world behavior. I wonder, though, if the same neural net would work across different pigs, or if it’s unique to the pig the neural network was trained with. Either way, you could probably use the same model and transfer the learning to a new neural net for a new pig (or even another animal?) much more quickly than starting from scratch.
The applications of this technology are vast, as demonstrated at the end of the presentation with the Q&A and various team members’ comments on why they are so interested in developing this technology.
Interestingly, it seems the primary purpose of the presentation was to raise awareness and attract qualified engineers to scale this device to mass production.
Also of interest is Elon’s primary purpose for starting Neuralink, which is to be able to compete with AI in the near future, which Elon currently considers to be the greatest danger to mankind. That is, computers can communicate much more quickly than humans can with each other, so in order to have a chance to fight computers that are communicating much more quickly, we as humans also need to communicate as fast. The human voice and typed words are extremely slow compared to how fast computers can communicate with each other. For example, it takes less than a second for computers to move data across the internet to the other side of the world, but humans would take a few seconds to type a sentence or speak over the phone to communicate a message. That’s just sending a message. The other component is comprehending the message quickly. Humans can’t do that on their own any more quickly than listening to a voice at 2x speed or speed reading as quickly as humanly possible. That’s where Neuralink would come in. You could communicate with each other by just thinking, and the person on the other end would understand instantaneously relative to how we currently consume communication.
All this is extremely exciting and intellectually stimulating, however another part of me is honestly a bit scared to see what happens with this. There will be both good and bad actors using these devices. As if our smart phones weren’t enough of a distraction in our lives, soon we will have yet another technology that will likely disrupt our lives for better or worse.
One question I have is, how would the device know for sure what you really want to do? For example, I could have a quick thought about doing something, but the device may interpret it to actually perform the action although the action was not something I intentionally wanted to do, or was unsure about doing. That is, sometimes I think about doing something, but don’t want to actually do it (I’m sure we all procrastinate). I wonder how much intent you need to have in order to trigger an action.
Regardless, congratulations to the Neuralink team! This is a monumental achievement, and I’m excited to see what benefits to society this technology can bring.