What is Neuralink? Neuralink explained: chip, experiments in pig Gertrude and monkey. How Neuralink works & vision of Elon Musk (cyborgs).
Elon Musk Neuralink
Halfway 2017 Elon Musk came in the news with Neuralink for the first time. The company wants to expand the human intelligence with artificial intelligence by implanting a chip in the head.
More specific: this company wants to be able to intervene on the neurons in the brain.
I have structured this article as follows:
- Video: How does Neuralink work?
- Vision of Neuralink (based on interviews from 2018);
- Other developments and initiatives;
- After space, Musk now enters the brain (based on a 2020 interview);
- Link in pig Gertrude (update after presentation in August 2020).
My video about Neuralink
How does Neuralink work?
On my YouTube channel I made a video about Neuralink: the technology, the demonstrations with pig Gertrude and monkey Pager, their ideas, criticism and my vision.
What is the vision of Neuralink?
Vision of Neuralink
Downloading a script for neurons that works for me in a specific way, copy this to someone else so that this person can do the same things as I can.. unfortunately it’s not that easy. The scenes from The Matrix, where Neo can download new skills, seem a long way off.
However, the perspective itself is interesting. On WBW [link at the bottom] is the nice comparison that we as humanity invented language about 50.000 years ago. That is a form of communication we still use. We added writing, printing and the internet.
But why can’t we communicate through our thoughts? That is the vision of Neuralink.
In September 2018 I talked to Elon Musk about this in the podcast of Joe Rogan:
Musk sees himself as a heavenly stormer, designing a whole new industry with his ideas, money and imagination. The goal of this industry is not only to make money, but to advance the whole humanity.
What are the advantages?
- Faster communication between people.
- More nuanced and accurate communication between people (according to Elon Musk himself).
- The points above are leading to better cooperation, more innovation, etc.
- Direct use of artificial intelligence. Think about directly understanding a foreign language.
Elon Musk is disturbed by ‘the output’ of the human brain. This is excruciatingly slow, compared to, for example, ‘the high bandwidth of the visual interface with the brain’ (the eyes) of people. In other words: data comes in quickly, but to talk or type? That process is really slow. His idea is to capture thoughts and convert them into text or movement. Then the humanity would communicate a lot more efficiently.
The output of the human brain is really slow.Elon Musk, Neuralink.
Can people then read your thoughts? Musk says that it’s not going to be that way. He thinks you have to make the choice to share a thought with someone else. Maybe we don’t even call it thoughts anymore. He can imagine that we will have a different word for the thoughts you keep to yourself and the thoughts you share with other people.
In addition, there are countless other challenges awaiting the scientists, researchers and technicians at Neuralink.
The tricky part, according to researcher Paul Merolla, is that the functioning of the neurons differs from person to person. A second problem is the countless connections neurons make with each other in a network. We do not yet have the computing power to map this out and intervene. A third problem is neuroplasticity. De connections in our brain change every moment.
This means there are a lot questions unanswered. How can you create enough bandwidth with those countless neurons and connections? How do you channel the energy to the chip? Can we chip all humans?
In an article for the Dutch newspaper Volkskrant, Esther Keymolen (University of Tilburg) indicates that you make people dependent of a commercial company: ‘You basically get a company in your head with this technology. It is already difficult to find out what technology companies do with all your data and how they make connections.’
You basically get a company in your head with this technology.Esther Keymolen, University of Tilburg
I also discuss this with neuroscientists Martijn van den Heuvel (Dutch Connectome Lab). “The question is whether we want that as society. Do we want that? Do we want an extra hard disk of our brain? Do we want to be influenced by a chip? You can say that you will lose a form of autonomy as a result.”
It also provides a paradigm the other way. Maybe then a real collective consciousness will arise. Then we really become a new species through the symbiosis of man and machine.
There is also criticism of the technology. Professor David Eagleman of Stanford University says that it’s more likely that we can genetically modify neurons. Rather than that we can analyse or interact with the brain with nanoscale robots. We do not necessarily need to introduce invasive technology in the brain for that.
In this part I will write about other interesting developments and initiatives about this subject.
In September 2016 I talked to Divya Chander during Singularity University the Netherlands Summit. She is Neural Data Scientist and anaesthetist at Stanford. In her research, the researchers try to analyse neurons and synapses very specifically. A downside of ‘deep brain stimulation’ is that it is an invasive technology. We bring electronics into our brain. There is a risk for other areas of the brain to be damaged.
Divya Chander (Stanford) thinks there are solutions for this problem. Tests are being done with light (optogenetic technology) and noise (sonogenetics) at mouses. Psychedelics are another solution. She fits this into her research, for example with ketamine. ‘Ketamine can help with opiate receptors and thus reduce addiction.’
Besides the example of Chander, there are other examples of developments in neurotechnology:
- Neural Dust
- Improve the memory
- Hearing colours
1. Neural Dust
Naura is a chip so big as a grain of wheat [link below]. The chip is controlled and driven by sound. The chips can (in theory) be used to study, monitor and control nerves and muscles (at a distance).
2. Prothesis in brain
A group of scientist at the University of Southern California (Los Angeles) placed a prosthesis in the brain of people who suffer from epilepsy. This prosthesis sends electric shocks to the hippocampus. This is the part of the brain where epilepsy and problems with memory are localised.
According to an article in de New Scientist the shocks were delivered per patient in a specific pattern [link below]. Here is the interesting part: a working memory test gave much better results than if the shocks were administered randomly.
3. Neil Harbisson
Neil Harbisson is a famous biohacker. He was born colour blind but he can hear colours with an implant. I often quote him in lectures on this subject. In my article about implantable technology, you can read a profile of Neil: Neil Harbisson.
I spoke with Neil and recorded our conversation on my YouTube Channel:
In 2020 Elon Musk visited Joe Rogan again to talk about Neuralink. I wrote a piece about this as part of my study in science journalism. For this I interviewed Dr. Mariska van Steensel (University of Utrecht).
After space, Musk now enters the brain
On May 7, 2020, Elon Musk joined Joe Ragon in his studio. Rogan is the host of an immensely popular podcast, the Joe Rogan show. The conversation starts about the name of Musk’s last child (X Æ A-12) and soon turns into Neuralink. A quote of the interview, which is viewed over 14 million times on YouTube. After more than 23 minutes, Musk says: ‘We are working on a chip that can cure all types of brain disorders.’
Take a look at the interview below:
From epilepsy to Alzheimer
Musk: ‘Think about epilepsy. We detect that and send a pulse to counteract the attack.’ He gives the impression that everything is possible with Neuralink: curing sight, regaining control of reluctant muscles, regaining speech after a heart attack and reviving memory in Alzheimer’s disease.
Neuralink works on brain-computer interfaces (BCI), sometimes called brain-machine interfaces (BMI). There are methods for human to communicate with computers through brain signals. At the UMC Utrecht Brain Center, Mariska van Steensel conducts research into brain-computer interfaces. She and her colleagues are working specifically on the Utrecht NeuroProthesis, a brain implant that allows a patient to use a speech computer with thoughts.
Hawking was still able to control a small muscle in his cheek.Mariska van Steensel, UMC Utrecht
The NeuroProthesis is for patients with severe paralysis, like the locked-in syndrome, which can be caused by a brain stem infarction (a stroke that does not get enough blood into the brain stem, PJ) – or an advanced stage of ALS. Due to their condition, they can hardly or no longer communicate. A well-known example is the late Stephen Hawking. Van Steensel: ‘Hawking was still able to control a small muscle in his cheek.’ For example, he could select letters on a computer, which converted the letters into speech.
The patients who van Steensel studies often can no longer use any muscle. The procedure is for a surgeon to drill small holes into the skull, place small electrodes on the brain and place an implant under the collarbone. That implant sends the brain signals from the electrodes wirelessly to a computer.
The first subject underwent this treatment five years ago. She communicates this way: letters light up one after the other on a matrix. By thinking on a specific action (such as clenching your hand) you make a so-called ‘brain click’. With that brain click you can select a letter or character. That way you can form a word or sentence letter by letter. ‘This is quite slow, but the first participant is really very happy with it.’
This is quite slow but the first participant is really very happy with it.Mariska van Steensel (UMC Utrecht)
The hardware from Utrecht differs from the technology that Neuralink uses. Neuralink places the electrodes in a different location. In a paper from July 2019, Elon Musk presents their approach. He describes it as a ‘BMI system with unprecedented high bandwidth’. Before this, the company used 96 thin flexibel wires, each with over 3.000 electrodes. A surgical robot can pierce six threads per minute through the skull into the brains.
Van Steensel: “They want to pierce a lot of small wires into the brain. We put the electrodes on the brain.” The conceptual advantage is that the data of Neuralink is much precise and detailed because the electrodes are smaller and can therefore measure smaller groups of brains. In a large group of brains, you have a greater chance that there are also cells that have a different function and that dilute the signal.
Still, Van Steensel cannot suppress her surprise when it comes to Neuralink’s promise to be able to remedy all neurological disorders as a panacea. Van Steensel thinks that is quite a bold statement: ‘Research groups all over the world have been working on brain-computer interfacing for years.’ In addition, you cannot put all brain disorders together. ‘You must have thorough knowledge of all those diseases. How it arises, what it looks like and what you need to fix.’
Testing on humans
In the interview with Joe Rogan, Musk is optimistic about the progress of the neural wires. He expects to perform the first experiments on humans within one to two years. Although the first tests with laboratory animals seem to be successful, the step to experiments on humans is not progressing so quickly. Van Steensel: “The regulations for testing on humans are strict. It seems hard to me to believe that they can test this on humans so quickly.”
She also sees a ray of hope. ‘The attention Musk gets generates a huge capital injection, which can accelerate the movements in our field.’ That is, if Musk has not find other interests in the meantime and starts living on Mars with his wife, X Æ A-12 and other children.
In August, 2020 Neuralink gave a presentation about the progress of their product.
Link in pig Gertrude
In August 2020 Neuralink organized a presentation. Due to the pandemic, it was broadcast via YouTube with a live stream. You will find the recording of the stream later in this article.
At the beginning of the presentation, Elon Musk spoke about the purpose of the presentation: to attract talented employees who want to work for Neuralink. ‘We don’t want to raise money today. No, today we mainly want to persuade people to work for Neuralink. That way we can further develop our product, making sure it is cheap and reliable so that more people want to wear it.’
The ambitions are still sky-high. Climbing mountains without fear. Playing a symphony inside your head. Discovering the nature of consciousness. Fix blindness, paralysis, deafness and mental illness. That is the promise Neuralink makes.
At the end it’s like a Fitbit, but inside your head with small wires in your brain.Elon Musk
During the presentation nobody mentioned a time frame or date on which this all would be possible. Elon Musk is also not known for meeting such deadlines, but still. With this kind of applications many people would be interested in Neuralink’s product. That’s the goal: ‘At the end it’s like a Fitbit, but inside your head with small wires in your brain.’
Before the presentation Musk announced online that he would show a spectacular demonstration. He first showed the new version of the Link. This is a disc with the diameter of a large Euro coin. The disc communicates with a telephone via Bluetooth. The Link contains 124 canals and has the battery life of one day. If people are going to wear this version ever, they will have to charge the Link every night.
After this, three pigs came on the screen. One of them didn’t have a Link and another had the implant removed. They both looked normal and healthy. The third pig (named Gertrude) had a Link inside her brain. The signals from Gertrude’s Link were visible on the screen. For example, the image changed when her snout contacted the ground, searching for food.
Watch the presentation of Neuralink in August 2020 below:
Science journalist Antonio Regalado from MIT Technology Review is not very impressed [link below]. Reading signals from the brains of animals is not very special for neurologists and neurotechnologists. He is especially critical at the promise Neuralink made to remedy a variety of cognitive disorders. The title of his summary of the presentation of Neuralink says it all. He describes the presentation as ‘an Elon Musk theatre show’.
In his presentation Musk refers to the rapid progress in the number of sensors on the Link and the possibilities to extract even more data from the brain, and vice versa. But Regalado writes: ‘You don’t necessarily solve conditions such as paralysis or depression with more computing power, but with more scientific knowledge about the electrochemical imbalance that these diseases cause.’
Despite the criticism from Regalado and other scientists, Musk sees an immense potential in Link and her successors. Think about summoning your Tesla with your thoughts and operating characters in a computer game like Starcraft with your brain. Over time Link should help you save and replay memories. He quickly adds: ‘That indeed sounds like an episode of Black Mirror. But yes, I think they are good at predicting the future.’
The first clinical trial will focus on individuals with spinal cord injury (lower body paralysis) or tetraplegia (four limb loss). The plan of this first trial is to test Link’s effectiveness and safety. If Neuralink and Musk further fail to deliver on their ambitious plans, it would be great if they could help patients with spinal cord injury of tetraplegia.
To conclude, what are other initiatives like Neuralink?
Musk isn’t the only one who believes in this, by the way. Bryan Johnson is the founder of a company called Kernel [link at the bottom]. In interviews, he frequently asserts that the question is not whether we are going to implant electrical chips in the brain, but when. Similarly to Neuralink, Kernel primarily focuses on the treatment of neurodegenerative diseases such as epilepsy, Parkinson’s and Alzheimer’s disease.
Bryan Johnson: ‘Intelligence is the most precious and powerful resource for humans. We’ve always built these tools, starting with the rock, thermostat, calculator. Now we have AI. Our tools and digital intelligence are increasing at great velocity. On the flip side, human intelligence is just about the same as it’s always been.’
Our technological tools and digital intelligence are developing at a faster and faster pace. On the other hand, our own biological intelligence has always remained more or less the sameBryan Johnson (Kernel)
Johnson and his colleagues at Kernel – including Theodore Berger, who worked there as a scientific advisor for a while – also see that there is still a long way to go. ‘The computations and algorithms carried in the brain are still largely a mystery to us.’ Another complication is the fact that the brain is constantly changing. Every second, a million connections are formed in the brain.
Musk’s idea may seem like a page out of a science-fiction story, but there’s some fascinating research taking place in the realm of brain-to-brain communication. Some leading publications in the past few years:
- In 2013, neuroscientist Miguel Nicolelis (Duke University) already demonstrated that he could connect rats to each other via a computer, using an implant in their motor cortex [link at the bottom];
- In the same year, neuroscientist Seung Schik Yoo (Harvard University) established that people with EEG electrodes on their skull could get rats to wiggle their tails [link at the bottom];
- One year later, professors Rao and Stocco (University of Washington) demonstrated that they could communicate with each other via EEG electrodes and a TMS headband [link at the bottom]. One simply thought of pressing a button, and the other actually pressed it after he received a magnetic pulse through the headband.
- In 2019, Cornell University researchers developed a Brain-to-Brain communication program, building on the aforementioned research by Rao and Stocco. The platform is called BrainNet.
Once again, we shouldn’t forget the fact that the brain is ever-changing and evolving. That means that, at least in the short term, downloading knowledge like in the movie The Matrix is not feasible or realistic. Memories aren’t just stored in one specific place in the brain, but are formed through associative connections between different brain regions.
The concept of brain-to-brain communication is also frequently explored in fiction, such as books, movies and TV shows: Neuromancer, Ghost in the Shell, Transcendence, The Matrix and Nexus are just a few examples. In almost none of these works, however, does brain-to-brain connectivity bring out the rosy utopian world described by Musk – although that may also just be typical of science fiction. I can personally relate to this as well: movies, TV shows and books that sketch some kind of gloomy dystopia of the future, do tend to hold my attention and fascination longer.
Do you want to know more about Neuralink?
Please contact me if you have any questions! Like if you want to invite me to give a lecture, presentation or webinar at your company, at your congress, symposium or meeting.
Or if you want to book a session with me as an expert consultant on this area.
This section contains additional information: my presentation on the subject, a few videos and the list with all links.
These are other relevant articles on my blog:
These are relevant books:
- Biography Elon Musk
These are external links that I have used, arranged per part.
- Website Neuralink
- Paper Neuralink
- Website Utrecht Neuroprosthesis
- Article Kernel
- Article Neural dust
- Article in MIT Technology Review (2020)
- Article in TechCrunch (2020)
How do you feel about the products and plans of Neuralink?
Please leave a comment!