Cyborgs: You too will be superhuman…

It was inevitable that at some point the development of Artificial Intelligence (AI), robotics and new materials science would one day cohere to enhance the human body in ways that were unimaginable just a decade ago. With advances in genetics, it is imaginable that humans and machines might one day be – well – joined at the hip.

We thought that this would be some way off in the future when cyborg astronauts (assisted by trusty old-fashioned robots, of course) might cope with intergalactic space travel. But it’s happening already, down here on Earth. Yes, some early basic cyborgs have already arrived – and will soon be knocking on investors’ doors.

Soon, you’ll be able to upgrade your eyes to attain bionic vision. You’ll be able to embed computer chips in your brain to enhance cognition. You’ll gad around in an exoskeleton that could give you Olympic athlete levels of performance in any sport in which you choose to participate…

In practice, however, this technology will first be exploited by the military before it becomes generally available to the likes of you and me. But it is just possible that you might live long enough to become superhuman.

Robotics meets AI

Elon Musk – of SpaceX and Tesla (NASDAQ:TSLA) fame – has had another brain baby. And this one really is going to change the world. It’s a company called Neuralink and its mission is to blend man and machine by means of brain-machine interfaces. (Tediously, the inevitable three-letter acronym is BMI.) Make no mistake, the cyborgs are coming.

Mr Musk’s motivation, as ever, is boundlessly ambitious. He wants nothing less than to save mankind. He is not alone in thinking that relatively soon artificial intelligence (AI) will overtake human intelligence and that that could be a problem. Will the intelligent machines be nice to us? If there is any possibility that the answer to this question is “No” then there is only one solution. We (the human race) must stay one step ahead of the machines by integrating their superior intelligence into our own. If you can’t beat them – join them!

The idea that BMIs can help humans accomplish basic tasks has been around for some time. The Swiss Federal Institute of Technology at Lausanne University has been working on a technology that enables tetraplegics to control their wheelchairs by the power of thought alone. Automotive manufacturers such as Nissan (TYO:7201) have been working on technology which will allow drivers to communicate with their smart cars (self-driving or not) by the power of thought.

Robots can already pick strawberries twice as fast as the most energetic human pickers. They are seen as essential to boosting productivity. Many people fear that robots will take jobs away from humans, though there are differences of opinion as to how many jobs are at risk. A recent report from the OECD estimates that only about 14 percent of jobs in developed countries could easily be automated. It envisages that robots will free up time for skilled employees by undertaking repetitive tasks and leaving more time for more creative activities. Another report by McKinsey estimated that 45 percent of all jobs could be completely automated using currently available technology and 60 percent could be partially automated.It estimated that the total number of jobs at risk globally from robotisation was 375 million.

A case in point is self-driving cars which will presumably put cab drivers and delivery drivers out of work in large numbers. But one can imagine that there will be an equal number of new jobs arising from maintaining and managing these vehicles as they zoom around our cities.

But – I know a lot of people won’t like this idea – if robots can do human things, what if humans could do robot things? Welcome to the weird world of cyborgs – and to the coming trend of cyborgation…

Exoskeletons and exosuits

Just as suits of armour enabled medieval knights to go into battle with enormous confidence and panache, robotic exoskeletons and exosuits can already transform ordinary mortals into augmented beings with extraordinary powers.Exosuits enable people to move faster, lift heavier weights and work longer hours than would ordinarily be possible. Check out a BBC video of a little boy with disabilities in Brazil using an exosuit to get around[i].

Cyborg technologies were much in evidence at the 2019 Las Vegas Consumer Electronics Show (CES) in January this year. One exhibitor was a San Francisco start-up called Roam which has developed an exoskeleton called Elevate for use by skiers. Basically, the attachments make your legs stronger. The company has secured $12 million in funding. Other exhibitors with equivalent products included Ekso and SuitX. Ekso’s devices have enabled paraplegics to walk again. But they also enable ordinary people to undertake tasks that they would not otherwise be able to do.


Prosthetic limbs have been around for a long time but there are also variants of these for people who are able-bodied. Mechanical trousers can help people with impaired movement to walk more easily. Mechanical vests can be worn to give, for example, warehouse staff or construction workers extra strength and endurance enabling them to carry out strenuous tasks for longer.

South Korea’s Lucky Goldstar (LG Corp. KRX:003550) is also active in this space with a home assistant called Cloi. Another Korean technology champion, Samsung Electronics (KRX:005930) also demonstrated its GEMS-H exoskeleton at the CES. These electric braces are worn over the hips and thighs and allow people to walk about 20 percent faster than otherwise. They also make climbing stairs almost effortless.

It’s not clear to me that these home assistants could yet be inserted into people’s heads – though I am quite sure that there is a laboratory, somewhere, working on that. Analysts at the Consumer Electronics Show estimated that the global market for human augmentation is currently quite small at about $600 million – but will grow to $3 billion by 2022.

Augmented Hearing

This is one better than noise-cancelling headphones by means of devices placed within the ear called (you’ve guessed it) hearables. All the big tech players are onto this. Amazon, Apple, Google and Microsoft are all working on devices which will enable the user to cancel out extraneous ambient noise and to concentrate on the desired sound – whether it be bird song, or a philosophy lecture.

When you walk around San Francisco these days every other person you see seems to be wearing Apple’s wireless headphone earbuds. These enable users to listen to music discreetly – and much more. Such in-ear devices can now be linked up to virtual assistants such as Apple’s Siri. Cynics might say that the tech giants could then sneak surreptitious ads into our ears, even while we sleep. But surely they wouldn’t do that…Would they?

Augmented Sight

One of Googles few product failures was their so-called smart spectacles. This product might have been innovative technology but the ergonomics were dire – and it just wasn’t cool. Nonetheless, the concept has been taken up by other players. North is now producing smart specs called Focals which actually look like a pair of reading glasses – now available for $999 in the USA. Using these, you could summon up the weather forecast or read an email while having a conversation with your boss – or even do something less innocuous. (As I write this I have just learnt that North is laying off employees.)

There are very few companies which are currently licensed to insert retinal implants in the USA but one of them is California’s Second Sight. In January it secured $40 million in a rights issue, the proceeds of which will be used to continue development of a bionic eye. Prototypes have already restored sight to blind people. Advanced bionic eyes might be able to widen the spectrum of light discernible to the human brain. If we could become dogs in our hearing range, we might yet become owls in our night vision.

Then there are augmented reality (AR) headsets. Proponents of AR claim that it is going to revolutionise the way we use media. The technology aims to make TV screens redundant by projecting images directly into our field of vision. So, for example, you could watch a heavyweight international boxing match in your own sitting room just as if you were at the ringside.

The Microsoft Hololens (a snitch at just $3,500) is one such headset – but there are competing products such as the Vuzix Blade smart glasses (a mere $1,099) and the Nreal Light mixed reality glasses. Magic Leap, a Florida start-up which makes headsets that can help (about which I have written before in the context of medical technology), amongst others, surgeons to carry out operations from remote locations, is already valued at an estimated $6 billion.

Augmented limbs

Currently, prosthetic limbs are a poor substitute for what nature endows us with – but that could be about to change. We can design bionic limbs but the real challenge is to replicate the operation of the nervous system and to facilitate communication between the prosthetic limb and the brain. As usual, it is a matter of both hardware and software.

A British prosthetics company called Open Bionics has already produced the Hero Arm. This is a 3-D printed arm which endows the recipient with huge strength. In January it raised £4.6 million in a funding round. The finance raised will fund its entry into the US market. In James Cameron’s new epic movie Alita: Battle Angel (a story about a bionic girl with robotic arms and superhuman martial arts skills), the 13-year old actress who plays the heroine, Tilly Lockey, is actually wearing Hero arms. Tilley lost her hands and forearms to meningitis when she was a baby but is now – quite spectacularly – fighting back, thanks to this extraordinary new technology.


A new branch of surgery called osseointegration can attach titanium studs to nerves located within muscles. This opens an entirely new field which has been called neuro-embodied design. Proponents of this technology think that it could be used, not just to heal the injured and the lame, but to further human potential in a future age of incredible athletic achievement.

Chip races

Amazon and Google are now vying to initiate a quantum improvement in microprocessor technology – computer chips to most of us. In so doing they are muscling in on a space previously dominated by such incumbent players as Intel (NASDAQ:INTC), Qualcomm (NASDAQ:QCOM) and Nvidia (NASDAQ:NVDA). And they are being eagerly followed by the likes of Facebook (NASDAQ:FB), Apple (NASDAQ:AAPL) andChina’s Alibaba (NYSE:BABA). Clearly, Amazon wants to advance the prospects of its Alexa voice-activated technology (which uses its proprietary Echo speakers). Similarly, Alphabet–Google (NASDAQ:GOOG) wants to develop further its Assistant device.

Hitherto, software companies have relied on hardware companies to develop systems that will run their software optimally. Now, in the early phase of robotisation and AI, software companies are backwardly integrating into hardware. Rather than coming up with a set of algorithms and asking, say, Intel, to implant them into a chip, Amazon is now designing its own chips and taking them directly to chip manufacturers like Taiwan’s TSMC (SHA:600751).

Amazon, of course, is no longer a purely retail business. Amazon Web Services (AWS) now has global clout in cloud computing and cybersecurity. AWS runs data centres which process data for hundreds of thousands of corporate clients. It is reported to control about 62 percent of the global cloud computer services market against Google’s 12 percent.

Equally, Apple has been building its own chips to power its smartphones since 2017. One such is the A12 Bionic chip which powers the iPhone XS. It has been reported that Apple plans to phase Intel out of making chips for its MacBook computers by 2020.

Google has been making its own Tensor Processing Units (TPU) since 2016 which are designed for machine learning tasks. These chips were used by Google subsidiary DeepMind’s machine that beat a champion Korean player in the notoriously mind-teasing game of GO!

Enhancing the human brain

While man’s capacity to develop new technology has produced remarkable results in the last 200 years or more – and even more spectacular results since the advent of digital technology only 40 years ago – the human brain itself, which uses that technology, has not really changed much in the last 300,000 years.

Our limb of the primate family tree – the hominins – split for that of chimpanzees and bonobos between six and seven million years ago in the Miocene geological era. Over several million years we hominins developed an upright, striding bipedal gait in order to explore the landscape more efficiently. By two million years ago we were using tools. The genus Homo sapiens (as we flatteringly call ourselves) first emerged about 300,000 years ago, the oldest fossil of which has been found at Jebel Irhoud in Morocco[ii].

So the brains that we use to contemplate the baffling abstractions of string theory or that we apply to build more efficient turbines are essentially the same apparatus that our ancestors used to stalk prey on the savannah or to build huts out of leaves in ancient rain forests. There is no suggestion that there is anything inherently wrong with such a brain – indeed the human brain has been described as the most complex structure in the known universe with an estimated 85 billion nerve cells with some 150 trillion connections. (Though even our amazing knowledge of physics and chemistry still does not explain how all this electro-chemical activity generates consciousness.)

But could the human brain be improved? In neurology at least, size is not everything. Dolphins and elephants have larger brains than we have but are not as intelligent (at least on our definition of intelligence: dolphins and elephants are much nicer than humans – but that is another conversation). In terms of pure computational power, human brains are quite limited – even the best arithmeticians cannot beat a humble calculator. And in terms of collating and sifting huge volumes of data, computers already win hands down. Human brains are particularly feeble in estimating probabilities (which is why – as I have often written in these pages – investors’ expectations are usually misplaced!).


That is why the idea is now current that the best way to enhance the human brain might be to develop effective ways in which it can interact with AI. We now understand that in the pursuit of AI we are not setting out to replicate the human brain – because, as brain surgeon and neurologist Henry Marsh has explained[iii], brains are nothing like computers. Computers stand alone as isolated entities but brains are attached to bodies without which they cannot function. Humans not only think, but also feel – we’re guided, for good or ill, more by emotion than calculation; and emotion, even if it arises in the brain, is always manifested in the body.

In fact it took a surprisingly long time for humans to identify the brain as the seat of intelligence at all. Aristotle thought the brain was a kind of radiator for cooling the blood. 500 years later, Galen thought the most important parts of the brain were the fluid cavities at its centre rather than the grey matter itself. Descartes saw the brain as a kind of hydraulic mechanism; late Victorian medics saw it as a telephone exchange. It’s only really with Freud in the early 20^thcentury that we have come to see the brain as the seat of both human reason and human personality. And nowadays, we tend to see it as a computer made of meat.

We know that the brain is plastic – that is to say that it can re-programme itself. If you blind-fold a subject, for example, areas of the cortex which normally process vision will start to process sounds within 48 hours. In computer-speak, the brain is constantly re-programming itself.

The idea of intelligent machines has been around for a long time. In Greek mythology there was the bronze automaton, Talos, which patrolled the island of Crete. But scientists have been thinking about artificial intelligence analytically since the great Alan Turing (1912-54) published his seminal paper On Computable Numbers back in 1936 – more than 80 years ago. Turing, by the way, was clear from the outset that the brain was not like a computer. It is possible, however, that AI will help us to understand how our brains work – and that, in turn, it could help us to develop better AI.

At a famous conference at Dartmouth College (New Hampshire) in 1956 the AI pioneer Marvin Minsky announced that machines would exceed humans in intelligence within a few decades. But progress has been much slower than Minsky foretold. Although computer science has leapt ahead in the last 60 years, AI has been a long time coming. But there has been a breakthrough in the last five years as a result of the development of neural networks and machine learning. These are concepts that have been around for decades but which have been finally realised by companies like DeepMind Technologies (now owned by Alphabet-Google) founded by Demis Hassibis.

Neural networks only resemble brain networks in a very loose way. They consist of layers of parallel programmes with complex feedback loops between each one. Each programme modifies its input to each other in accordance with predetermined parameters. In this way the entire system can “learn” from experience. They still do not understand what is going on and cannot explain things – they are no nearer to being conscious than were Turing’s early valve-powered computers. It is therefore much too early to talk about computers that can “think”. Google Translate works (up to a point) not because it can “speak” languages but because it has trawled the entire internet to match an English phrase with (say) a Russian equivalent.

So far all AI programmes can only perform one task. This form of intelligence, as Henry Marsh points out, is reminiscent of some of the patients described in the writings of the late Oliver Sacks (e.g. The Man who mistook his Wife for a Hat). They are like “people” who can perform remarkable feats of calculation and yet are hopeless in everyday life.

The ultimate goal of AI developers is not AI but general artificial intelligence (GAI). That is a computer system that can mimic a range of human cognitive behaviours and therefore imitate human intelligence. Just as neural networks were inspired by brains, so now there is interest in neuromorphic chips. These are computer chips which are designed to resemble nerve cells. This could make computers much less energy-intensive.

The Human Brain Project (HBP), a pan-European research programme, started out with the vaulting ambition of building a brain using computers. Whether this ambition is realistic is highly contentious. One part of the HBP which uses neuromorphic design is the SpiNNker computer at Manchester University. Another research team at the Oxford Computational Neuroscience Lab recently published a paper arguing that a better understanding of how the brain works will holds the key to advances in AI.

AI could one day be harnessed to upgrade our brains – either through implants or by connecting our brains to remote supercomputers. But that is not in view yet. There is no risk that intelligent computers are going to enslave us in the near future (as we have enslaved animals for millennia).

Cyborg soldiers and astronauts

The first fully paid up cyborgs will probably first appear on a battlefield coming your way soon…Seriously, soldiers in the world’s leading armies are already kitted out with an incredible technological array in their helmets from infra-red vision goggles to GPS systems and laser-guided weaponry.

The Russians are very active in this field – check out a video on YouTube of President Putin inspecting a robot warrior. Interestingly, last year, Mr Putin made a speech about the ethical implications of using genetics to breed an army of soldiers without any sense of mercy. This may have been for the benefit of the Orthodox Church, prelates of which were present and of which he is a practicing member. But it also suggests that this is the kind of theme in play in Russian defence circles. (If you want to know what is going on in Russia it pays to read those of the President’s speeches which are in the public domain. I have always thought that spies are barking up the wrong tree: true intelligence (as in investment) lies right under our noses.)


In the long-term future cyborgs will make much better astronauts than humans as they could adapt more effectively to persistent zero-gravity and exposure to radiation. They may even be up to the challenge of the first mission to our neighbouring stars in search of Goldilocks-zone exoplanets which might support life…

If there isn’t one already, there will probably be a resistance movement against the rise of cyborgs, possibly even one rooted in religious belief. Although, to the extent that we already have prosthetic limbs or a pacemaker in our hearts or a chip measuring our blood pressure (soon to be quite normal) the process of cyborgation has already begun.

Mr Musk’s extraordinary vision and drive may even inspire a colony on Mars by the end of his lifetime (hopefully, he should live until about 2065 if he doesn’t overdo the weed). But there may be one thing Mr Musk hadn’t thought of when he established Neuralink. What if the cyborgs take a dim view of us mere non-cyborg humans and decide we are no longer required?

Action

The advancement of AI and cyborgation (as I call it) involve the development of multiple layers of technology combined in innovative and visionary new ways. The niche players which are working on exosuits and prosthetics are at the venture capital phase and are not easily investible. We can assume that many of them will be acquired in due course by the technology behemoths. It was probably inevitable that, in researching this article, the usual suspects came up again and again: especially Apple, AmazonandGoogle.

Last year I advocated that the social media component of the FAANGS – namely Facebook (plus poor old Twitter) – should be disaggregated from the popular acronym. (Though AANGS sounds like something you could catch.) The big three are at the forefront of numerous components of AI – though don’t underestimate Facebook’s competitive advantage in facial recognition technology (and possibly in voice recognition as well).

Overall, despite the siren calls predicting the downfall of technology stocks, in my view investors should keep a significant part of their long-term portfolio (depending on its risk profile and size) allocated to one or more high quality diversified technology funds. (I have mentioned the Janus Henderson Global Technology Fund in this context before). I also think AI/cyborg technology will impact the defence sector very soon – and I’ll have more to say on that shortly. For now, I’d just mention Lockheed Martin Corp. (NYSE:LMT) and Northrop Grumman Corp. (NYSE:NOC).

[i]Available at: https://www.bbc.com/reel/video/p06tynw5/the-boy-with-the-robot-body?ocid=gnl.display.house-banner.bbc-gnl..BBCReel

[ii]See Evolved to Exercise, by Herman Pontzer, Scientific American, January 2019.

[iii]Henry Marsh is author of Do No Harm: Stories of Life, Death and Brain Surgery and A Life in Brain Surgery. I have also drawn on his article Can we ever build a mind? Financial Times, 16 January 2019.