Earlier this year, Elon Musk announced a brain-computer venture called Neuralink whose end goal is to fuse of our brains with machines. Only Musk can start a venture that makes colonising Mars look like the easier challenge.

It’s true. Neuralink may be the most complex and difficult to achieve of Elon’s ambitions primarily because its seeks a solve for one of the biggest hurdles for the human brain —  understanding itself. This week’s #FactorFuture peers beneath the skull of Silicon Valley’s latest obsession.

The ghost in our shell

Interfacing brains with machinery isn’t a radically new idea. For the last four decades we’ve been using electronic devices to restore functionality to parts of our nervous system and sense organs under the umbrella of “neuroprosthetics”. Cochlear implants, for instance, collect audio from the world and pass them on directly into the auditory nerve simulating the cochlea. More than 350,000 people use it around the world today. Bionic eyes have begun enabling rudimentary vision.

Neuroprosthetics, however, has focused on replacing missing biological function by connecting the nervous system to electronic devices. Musk, on the other hand, wants to connect the human brain (or the central nervous system) directly to a digital superhighway. He calls these “neural lace” — a term he is borrowing from the science fiction of Iain M Banks.

In Banks’s Culture series, all the beings living in a vast multi-star system are implanted with bio-mechanical implants (called neural lace) which meshes with the brain and enables everything from communication to storage and information retrieval to VR-like display to machine interfacing. Laces control gland secretions and fill their hosts (the beings having them) with a rush of adrenaline when they need to wake up in the morning. They also allow machines to be controlled as a natural extension of the body.

This is a popular theme sprinkled across cyberpunk fiction. From Neuromancer to Ghost in the Shell to The Matrix (heavily inspired by Ghost in the Shell), cybernetic enhancement (especially brain augmentation) often provides the plot complications and conflict of ideas.

These stories also provide a blueprint of what’s possible. Human intelligence, when released from the constraining bandwidth of our ability to communicate with other humans and machines, is capable of getting powerful by orders of magnitude.  It is towards this vision that Musk (and few others from the silicon valley) seem to be diverting interest and money in millions of dollars.

Brain computer interfaces

While the idea of enhancing human intelligence by interfacing with machines is getting more mainstream now, it has been around for a while.

Musk’s “neural lace” is more commonly referred to as the brain computer interface (BCI), a term coined back in 1970 in a paper published by University of California, LA (UCLA) following a contract from the Defence Advanced Research Projects Agency (DARPA). DARPA itself has a long, secretive history of working on programs to take forward BCI.  Since 1974, it began investing in a biocybernetics programme intended to investigate brain waves for both communication and monitoring the physical and emotional state of subjects.

There have been a few early successes using BCI on animals. Monkeys have controlled robotic limbs and navigated robotic wheelchairs using just their brain. Theodore Berger, a professor of biomedical engineering in university of California, proved back in 2002 that it was possible to simulate the hippocampus using math and algorithms (hippocampus is the part of the brain that helps remember things). In 2008, scientists in Japan were able to reconstruct images from the brain in black and white with a 10×10 resolution. In 2014, a bunch of neuroscientists from around the world demonstrated that it was possible to transmit thoughts through just the brain waves between two humans.

But it’s the entry of the silicon valley entrepreneurs that’s bringing all this into the mainstream conversation.

Over the course of last year, two private commercial ventures have begun working on BCI —  Elon Musk’s Neuralink and Bryan Johnson’s Kernel.

Neuralink’s plans and current activities are largely shrouded in mystery. It’s mission statement talks about “developing ultra high-bandwidth brain-machine interfaces to connect humans and computers” and its website lists down a range of positions from polymer scientists to biomedical engineers to senior scientists, brain-machine interfacing. It stays away from words like neural lace and although Musk’s eventual goal is “human enhancement”, Neuralink is likely to be working on brain diseases in the short term.

Kernel, founded by entrepreneur Bryan Johnson with $100 million of his own money, is attempting a similar feat . As the first step, it too, is looking for medical breakthroughs that solve major neurological problems.

But both these entrepreneurs have made no secret of their desire to eventually build a world where the human brain is faster, more powerful and can create amazing things. It’s a world view incorporating machines as part of the humanity rather than letting one evolve rapidly on its own.

Read: I hope Elon Musk and Bryan Johnson can accelerate the path to brain implants: Ramez Naam

The other Silicon Valley overlord to join the fray recently was Mark Zuckerberg, who wants to do away with slow typing fingers and slightly less slower speech and instead get your next status update directly from your brain. In this year’s F8, Facebook’s annual developer conference, Zuckerberg revealed that about 60 engineers were working on using non-invasive methods like optical imaging to detect the words in people’s brains and translate them into text. Facebook’s forays in this space are headed by Regina Dugan, a DARPA alumni with experience working on BCI.

Telepathy-like communication is one of the dream applications (for a platform like Facebook) of BCI along with connecting the brain directly to the cloud, downloadable learning and control of machines with thought.

Even the chip makers are getting in on the act. ARM, a US-based semiconductor company, recently announced that it is looking to design ultra-small, ultra-low power chips that could potentially go inside human heads. While all this sounds exciting (and scary), it begs the question of how close are we to having any of this become a reality.

Cracking the neural code

A huge list of scientific challenges lie along the way towards a fully evolved BCI. The fictional ‘cyborg’ goal will likely take several decades and cost billions of dollars in research (the kind of foundational research that needs to go in before it can even be brought to market).

Read: From glassholes to cyborgs: Wearables look to redeem themselves in their second coming

Even with the various demonstrations around using brain waves to control movement, offer vision or enable sounds, the progress so far has been limited and stuttering: Slow, limited or imprecise motor control, low-resolution vision and low fidelity hearing.

More importantly, we need invasive surgery today to plant electrodes and sensing equipment in the brain in order to read and respond to the impulses.  This is a high-risk procedure today  —  one that neuroscientists feel will take a long time to overcome. We aren’t going to happily let scientists break open our skulls and plug into our grey matter any time soon unless we are really desperate.

Johnson himself admits that currently there is no technology that allows you to truly understand and read the brain waves without invasive procedures. This is the reason Facebook is working on non-invasive ways to achieve the same. Researchers at UC Berkeley are working on neural dust  —  millimetre-sized sensors that could go into the human brain in thousands and report on what’s going on there.

All of this goes back to one overarching point: We’re still a long way from understanding how the billions of neurons in our brain work together to provide the sensations and thinking capable in human beings. We are yet to fully understand how the brain keeps learning new things (neural plasticity) and how that’s often unique for each individual.

Not that we’ve not been trying to break down the brain.

In 2013, the European Commission awarded Henry Markham (a neuroscientist) $1.3 billion to building a simulation of the human brain, a feat that he promised can be achieved in a decade ago in 2009. Four years hence and even as we approach the decade mark that Markham promised (2019), the Human Brain Project has been plagued by disputes and is yet to drum up additional investments. In 2013, the US launched the BRAIN initiative, a billion-dollar moonshot, to understand and map the human brain and accelerate brain-inspired computing. Countries around the world have their own programmes to study and understand the brain.

But it’s a long hard grind.

There are about 100 billion neurons in the human brain (just a little less than the number of stars in our galaxy) and about a 100 trillion connections (less precisely known). These work using a combination of electro-chemical impulses creating millions of networks and pathways (new ones are being created all the time — an activity we call learning). It’s safe to say that this is one of the most massive and complex systems humanity is trying to understand currently.

Yet, we’re still fighting with limited tools. We often get to observe only a subsection of these in very few people (often with brain related health conditions). We do not have ways to look at real time brain activity at the speed of thought in human beings. We don’t have microscopic resolutions to dive deeper.

In order to build these wonderful brain interfaces, we need to truly understand the brain and in order to do that, we need to collect enormous amounts of data from it. Today, we are still trying to solve for how to best do it.

The mind is what the brain does

Marvin Minsky, considered the “father of AI”, famously proposed the view that there is nothing called our “mind” or perhaps even ourselves beyond the activities (the electrical and neural) that take place in it. It is disconcerting to think that the unique ‘me’, that each of us rate so highly, is probably all just a bunch of black-grey goo in our heads. And the body is just packaging?

No wonder then that it shocks us to imagine a future where we could potentially manipulate this goo and achieve symbiosis with machines. It surfaces the uncomfortable existential question of ‘what makes us human?’.

The proponents of BCI believe that this is the only way forward. Ray Kurzweil, the poster boy for the trans-humanist approach to the future, thinks that it is an inevitable step in our evolution. In fact, according to him, we’ll likely transcend what we now know as humanity and become a new type of being, not much unlike how Neanderthals got replaced by human beings, except in this case technology would have accelerated the evolution.

Kurzweil believes that by 2030 we’d have nanobots in our brain connecting it to the cloud and providing full immersion VR. Even though this may seem too far-fetched, Kurzweil’s statements have grown in stature as more of his predictions have come true with every passing year.

The entrepreneurs working on BCI often bring up the doomsday fear of AI’s increasing capabilities and what would happen if human intelligence gets ‘left out’. In this framework of thinking, enhancing the power of human intelligence is being seen as the only way to stay relevant in this world. It pits an even more scarier, dangerous concept to make BCI look less threatening (perhaps a smart way to change the value proposition).

Toying with the brain brings about two types of fears in itself. The first is the whole host of ethical and moral issues could arise from being able to manipulate thoughts, moods and actions. As advances are made in BCI over the coming years, questions around neuroethical guidelines for testing and real life augmentation start surfacing. We’ll have numerous legal and regulatory questions around whether to allow augmentation that does not solve a medical issue and how do we treat the augmented individuals as part of the society? Then there are likely to be questions around social equality and affordability of such procedures as well.

The second is a slightly far-out, cyberpunk-driven fear like brains being hacked and mass control. Again, we can look to science fiction for a laundry list if we ever need one. Characters in Iain M Banks’ Culture books often seek liberation from the ‘neural lace’  —  it turns out losing our agency isn’t such a fun thing after all. But that could be the least of the worries. A connected brain could open us up for a range of novel dangers from simulated torture, loss of privacy, brain hacking and simulated control (like in The Matrix).

But perhaps it’s too early to call on the dystopia of brain-computer interfaces. Advances made here will have a meaningful impact in treating a range of disabilities and diseases, helping millions of people live better lives with little to no compromise. More importantly, we are still at the beginning of the journey in understanding how the human brain works. In that sense, it is the new ‘rocket science’ and brain is the new space. It needs a lot of hard research and time before we begin to worry about its sinister applications.

---