A Blog by Jonathan Low


May 20, 2017

The Battle For Your Mind: Could Elon Musk Beat Mark Zuckerberg?

Facebook's technology is non-invasive but entirely new. Neuralink, Musk's company, is trying to goose existing science. Experts tend to believe Musk's approach will reach its goal first.

But then both of them will have to convince governments - and humans - to participate. JL

Dyani Sabin reports in Inverse:

Facebook wants to create a system that allows people to turn their thoughts into text in a computer at a rate of 100 words per minute by 2019. Neuralink (Musk's initiative) is essentially trying to solve an engineering problem with the nascent brain computer interface technology that we already have.
In the next few years, you might be able to text your friends about the annoying person in line with you at the post office without taking out your phone. Instead, you could kvetch straight from your brain, using a device created by Facebook or Elon Musk’s Neuralink, that connects your mind to the cloud.
And if secretly kvetching isn’t your style, Facebook’s brain-to-text tech has the potential to create an entirely new language, and Neuralink is Musk’s attempt to get humans to survive the A.I. uprising. Either way, the goal is to upgrade communication and thought to a new level.
While both companies have similar end goals, at least one scientist from Neuralink seems to have a better chance of making this happen sooner.
“Neuralink is developing an implantable device to read the brain’s electrical signals, which, in my opinion, is more technically promising than Facebook’s quasi-ballistic photon project,” Junjie Yao, a biomedical engineer at Duke University, tells Inverse.
In the long run, Facebook is trying to develop an entirely new method of seeing what’s happening inside the brain. It’s an ambitious approach that has the potential to change our understanding of how physics works. Neuralink is also undertaking an ambitious task, but instead of re-inventing physics, it faces a challenge of scaling up our existing medical and engineering technology. It’s a difficult problem, but Yao thinks it makes Neuralink more likely to achieve its goals in the near future. The final products have the potential to change the way we communicate, and the way we interact with the world.

Facebook’s 2019 Prediction

Facebook wants to create a system that allows people to turn their thoughts into text in a computer at a rate of 100 words per minute by 2019. This was announced by Regina Dugan, the former DARPA program manager who is now the head of Facebook’s Building 8 Development Team, in her presentation at the F8 Facebook Developer’s Conference on April 19.
“It sounds impossible but it’s closer than you may realize, and it’s just the kind of fluid, human-computer interface needed for [augmented reality],” Dugan said. “Even something as simple as a ‘yes-no’ brainclick, would fundamentally change our capability. A brain mouse for A.R.”
The key to the tech is that it will be totally non-invasive, work in real time, and on the level of individual neurons. So Facebook is attempting to read your mind without injecting micro-chips into your jugular, a method Musk has talked about for neural lace. “I think it’s not totally impossible if Facebook has the right technology,” says Yao. Building that technology is likely going to take more than two years, because it will probably require developing a different understanding of the physics of how light moves through the brain, says Yao.

“100 Words Per Minute”

When asked about the differences in the technologies, a Facebook rep told Inverse in an email: “We can only discuss what Facebook is working on: A system capable of typing 100 words per minute, straight from the speech center of your brain – this is 5x faster than you can type on your smartphone today. We will do this with non-invasive, wearable sensors that can be shipped at scale.”
The non-invasive element is what allows it to be scalable, and Facebook has already outlined that it plans to use a combination of ballistic and quasi-ballistic neurons to achieve this. It’s not yet clear exactly how it will work, but the leading edge of brain-imaging technology has a few clues, including super-sensitive photon detectors, or adjusting the paths of each photon to increase the signal.
“If it was totally non-invasive that would be revolutionary,” says Yao. “If they do minimally-invasive, that’s a whole different story.”
The different story is Neuralink, which is essentially trying to solve an engineering problem with the nascent brain computer interface technology that we already have, says Yao. It is more likely to happen, but is not as exciting on a research level.


“Your phone and your computer are extensions of you, but the interface is through finger movements or speech, which are very slow,” Musk told Vanity Fair in March. “For a meaningful partial-brain interface, I think we’re roughly four or five years away.”
Despite his regular comments on neural lace, it’s not heralded as the most popular option for a brain-machine interface. In fact, the WaitButWhy post, and Musk himself, hasn’t highlighted a specific technology that Neuralink will pursue. Here are three ideas potentially in play at Neuralink:
  • Neural lace is a structure that could be inserted into the brain through the jugular that unfolds into a net-like series of superfine electrodes that interface with individual neurons.
  • Neural dust. It’s similar to neural lace but is a network of nano-electrodes that fuse to the brain, coating it like dust instead of surrounding it in a net.
  • -Micro-brain computer interfaces, such as computer interfaces attached to silk, electrode arrays printed on the brain, or micro-electrode implants inserted into the brain.
 There isn’t yet a brain-computer interface that works in humans without cutting open the skull and looking at a few very specific neurons, says Yao. “But with all that said, it’s a lot more promising than the quasi-ballistic photon technology, which is not actually in existence right now.” Both products have the potential to significantly alter our relationship with technology and our concept of the human mind. Although both products sound kind of crazy, technically both are possible, even though experts have questioned the aggressive timelines the two teams have suggested for each product. Either way, Ray Kurzweil, a leading futurist, has predicted that the moment when computer intelligence and human intelligence merges, called the singularity, will happen in 2045. And Facebook and Neuralink are getting us closer to that future.


Post a Comment