Advances in brain-computer interfaces and prosthetics could improve the quality of life for millions, but they carry risks that even touch upon eugenics.
“Steve Austin. Astronaut. A man barely alive. Gentlemen, we can rebuild him. We have the technology. We have the capability to make the world’s first bionic man. Steve Austin will be that man. Better than he was before. Better, stronger, faster.”
If you were a kid during the 1970s, ‘The Six Million Dollar Man’ was peak TV and, arguably, James Bond’s closest cultural competitor. Virtually everyone could recite its opening narration. But while 007’s main physiological attribute was an indestructible liver, Steve Austin was cybernetically enhanced with two superstrength legs and an arm, and a computerised eye.
His creators were inspired by work being done even then towards increasingly capable prostheses for severely wounded veterans and other amputees (the fictional Austin was a horrifically injured USAF test pilot) and the prospect that artificial limbs could be enhanced with electronics. After all, the first programmable production line robot, Unimate, entered service at General Motors in 1961, 12 years before Austin arrived on screen.
It is only now that we are approaching the reality imagined in Martin Caidin’s original novel ‘Cyborg’ and the retitled series.
During 2023, the US Food and Drug Administration (FDA) is expected to approve and monitor small-scale clinical trials of at least three rival brain-computer interface (BCI) systems.
These systems seek to read signals from neurons than can be used to provide subtle control of and ultimately feedback from artificial limbs, to stimulate nerve passages to restore movement after spinal damage, and, where neither of those options is viable, to give patients control of digital tools.
The BCIs on the launch pad are Synchron, partly backed by the investment funds of Bill Gates and Jeff Bezos, with trials already under way; Neuralink, backed by Elon Musk; and Paradromics, with backers including the US National Institutes of Health and Department of Defence (DoD), and family of the late Ray Dolby.
They are not alone. Meta and Alphabet have BCIs in the lab. Leading entrepreneur Peter Thiel is an investor in Blackrock Neurotech, a spin-out from pioneering research at the University of Utah. China has made BCIs priorities under plans such as Made in China 2025 and its AI strategy. Both the EU and UK have research programmes. Meanwhile, much of what is emerging has grown from seed-funding by the DoD’s Defence Advanced Research Projects Agency (Darpa) dating back to the beginning of the century and continuing.
Big hitters. Big money. Never mind ‘The Six Million Dollar Man’; the BCI market is forecast to reach $6bn within a decade, says Future Market Insights.
BCIs support a persuasive vision: restoring human function and dignity to millions who suffer and, in the case of veterans, have made huge sacrifices. But it is becoming a controversial one, too.
The same branch of speculative science fiction that gave us Steve Austin has produced his antitheses in, most famously, the Daleks and Cybermen from ‘Doctor Who’ and the Borg from ‘Star Trek’. They warn of digital and prosthetic augmentation to living creatures progressively causing their evolution into drone-like, fascistic supervillains.
Concerns around the latest breakthroughs in neurotechnology do not go that far, but still need addressing. Could BCIs be perverted to foster eugenics? Could they turn digital divides into chasms by creating a species of ‘post-human’ overlord? Could they combine with AI in unanticipated and dangerous ways?
Arguments and ethical debates previously associated with biotechnology, particularly around Crispr gene editing, are spreading into electronics.
The main brake on abuse right now comes from regulatory agencies like the FDA constraining today’s wave of innovation to justifiable medical applications. Many of the companies involved have also committed not to allow their products to have any military use other than towards the care of veterans – no supersoldiers (well, not from them, at any rate).
However, even here some critics point out that while the basics of what we today consider plastic surgery were developed to aid gravely injured servicemen and women during two world wars last century, they eventually became available to those who could afford them as elective enhancements.
A debate like the one we are seeing around AI in its independent form is intensifying in parallel around BCIs. We need to start with the enabling technologies. Advances there are the fuel for that debate.
Anyone who has seen YouTube clips from Boston Dynamics knows that the mechanical and control capabilities of robotic limbs and torsos has reached an extraordinary standard. Some of that sophistication has been making its way into human-worn prostheses.
The LUKE Arm is officially named for ‘life under kinetic evolution’ but really after Luke Skywalker’s robotic hand from ‘The Empire Strikes Back’. It was developed by DEKA Research, the company of Segway and insulin-pump innovator Dean Kamen, with funding from Darpa. LUKE has been offered to US veterans since 2017 and is marketed more widely by Mobius Bionics.
The prosthesis has three configurations: radial (hand), humeral (forearm) and shoulder. The company claims it is the only one to offer the third of those. It has up to 10 powered joints and allows the wearer to vary direction and intensity, including grip.
There are several control options, including electromyographic sensing of nerve activity. However, reflecting the current limitations of that technique, there are physical alternatives. These include a pressure switch or transducer and a foot-worn control.
In 2019, researchers at the University of Utah gave LUKE an upgrade. By implanting a 100-microelectrode array directly on a user’s nerves and linking it to an external computer, test subjects were able to not only control the arm but also sense what it was doing. “It almost put me to tears,” says trial participant Keven Walgamott. “It was really amazing. I never thought I would be able to feel in that hand again.”
LUKEs cost $100,000 each, though. The additional cost of the Utah Slanted Electrode Array and associated infrastructure put the package beyond most potential users due to the care ceilings imposed by insurers and the limited financial resources of public healthcare providers like the NHS.
Researchers at Johns Hopkins University’s Applied Physics Laboratory conducted similar research, also supported by Darpa, at around the same time on their own Modular Prosthetic Limb, using a direct nerve connection for control and feedback. That work is yet to be commercialised.
Image credit: Alamy
As brilliant as all these projects are, from the mechanical to the digital, they illustrate why BCIs are considered a necessary next step, not only for greater sophistication and more robust connections but also economies of scale.
BCI research has been going on for decades. Previous implementations have included caps packed with sensors (which encounter signal-to-noise and resolution issues from trying to read data through the skull) and so-called Utah Arrays of intracortical implants (which involve a comparatively invasive surgical procedure – these are Blackrock Neurotech’s original technology).
One major challenge has been to reduce the cost of surgery by using a less invasive technique. Today’s preferred approach leverages miniaturisation so that a BCI can be inserted as a stent and use blood vessels as sensor highways, taking advantage of the fact that those in the brain are wider than elsewhere in the body. Stent insertion is a well understood process.
Synchron implanted its first Stentrode in a US patient in July 2022, after earlier tests in Australia. According to the surgeon who performed the endovascular operation, it was minimally invasive. “The implantation procedure went extremely well, and the patient was able to go home 48 hours after the surgery,” adds Dr Shahram Majidi of Mount Sinai Hospital.
Synchron CEO Dr Tom Oxley captured the overarching objective: “Our technology is for the millions of people who have lost the ability to use their hands to control digital devices. We’re excited to advance a scalable BCI solution to market, one that has the potential to transform so many lives.”
Oxley and his rivals aim to develop BCIs that address multiple conditions compared to the approaches taken by Utah and Johns Hopkins Universities, which can be seen as more bespoke.
The focus of Synchron’s ongoing FDA trials is on enabling severely paralysed patients to use point-and-click communication devices (smartphones, laptops and OS-neutral). But that could extend to controlling prostheses and generating neurostimulus, all based on a general-purpose BCI produced in volume.
To get some idea of how far BCIs could soon go, researchers led by Switzerland’s École Polytechnique Fédérale de Lausanne (EPFL) recently reported on a trial that used one to enable a patient to start walking again after being paralysed by a spinal injury.
“This brain-spine interface (BSI) consists of fully implanted recording and stimulation systems that establish a direct link between cortical signals and the analogue modulation of epidural electrical stimulation targeting the spinal cord regions involved in the production of walking,” the EPFL team explains.
“A highly reliable BSI is calibrated within a few minutes. This reliability has remained stable over one year, including during independent use at home. The participant reports that the BSI enables natural control over the movements of his legs to stand, walk, climb stairs and even traverse complex terrains.”
EPFL has meanwhile developed its own small-scale cortical implant that can be inserted between the brain and skull through a 2cm hole. It expands six spiral arms over a 4cm diameter; the arms’ shape maximises contact between electrodes and the cortex. The implant is thin enough to sit in a 1mm gap between the skull and the brain’s surface, minimising the risk of tissue damage.
BCI technology is arguably on a recognisable path for silicon economics. Multiple deep-pocketed players are promoting competing implementations based on general-purpose devices in advanced packaging with software differentiation – with many on the road to commercialisation. Joining existing private-sector players, EPFL is transferring its array to a spin-out, Neurosoft Bioelectronics. But this rate of progress is raising some flags.
One issue that has dogged prostheses has been their affordability in the developing world. Countries there have more agrarian economies and, as a result of globalisation, have often become manufacturing hubs. This makes industrial accidents statistically more common, according to groups such as IndustriAll, a global alliance of trade unions. These countries are also more vulnerable to conflict.
Technology is helping. The open-source e-NABLE group has 40,000 volunteers in more than 100 countries who collaborate to create upper limbs for those in need with 3D printers. Researchers at De Montfort University in Leicester developed a way to produce limb sockets from recycled plastic at a unit cost of £10, against a previous benchmark of £5,000.
Yet as the best aids benefit from greater electronic integration, the digital divide once more raises its head. Concerns here are not confined to cost, availability and functionality, although all are vitally important. Others have been fuelled by Elon Musk and his stated objectives for Neuralink.
“We are already a cyborg. It’s just that you have a digital version of yourself online in the form of your emails, your social media and all the things that you do. And you have basically superpowers with your computer and your phone and the applications that are there. You have more power than the President of the United States had 20 years ago,” Musk told 2016’s Code Conference.
“But the constraint is input/output. We’re I/O-bound, particularly output-bound. Your output level is so low, particularly on a phone. Your two thumbs tapping away – this is ridiculously slow. Your input is much better because we have a high-bandwidth visual interface – the brain; our eyes take in a lot of data.”
The answer for Musk concerns how we relate to AI – and avoid being made pets or even destroyed by it. He foresees the need for a “neural lace” in the form of a digital layer above the cortex that would allow “merging in a symbiotic way with digital intelligence”.
Neuralink’s existence became public a few months after his comments, but they straightaway set alarm bells ringing around BCIs, particularly in how his proposal appeared to smack of transhumanism.
Originally articulated in the 1950s by Sir Julian Huxley, brother of ‘Brave New World’ author Aldous Huxley, transhumanism argues that mankind has a responsibility to use its mastery of technology to accelerate beyond natural evolution and create better people and better societies. Sounds good, but critics feel this utopian vision merely tries to give eugenics an acceptable face.
As viewed in terms of its potential impact today, critics view renewed interest in transhumanism as raising the danger of stratifying humans and ‘post-humans’, much as Aldous Huxley’s novel did into Alpha, Beta, Gamma, Delta and Epsilon castes. This new, more insidious digital divide is also discussed in terms of the predicted merging of humanity and machine at The Singularity.
More recently, political scientist Francis Fukuyama summarised the social objections being levelled at transhumanism by calling it “the world’s most dangerous idea”.
BCI researchers get seriously antsy whenever this is brought up. Their focus, they say, is medical and on making life better for those whose quality of life is far below that of most of us, not on extending a welcome to our new cyborg overlords. Transhumanism is, if anything, an unwelcome distraction.
The charges against contemporary transhumanism may not be entirely fair (though that is a tough one). Musk would likely argue his stance is existential rather than philosophical. But the concept sits alongside other controversial ones that include effective altruism and long-termism.
All are seen as having much wider support in the technology community than is recognised and as raising societal challenges which that community seems unwilling to address in ways that civil society can understand.
The likely combination of BCIs with AI compounds those concerns. On one level, broad-based BCIs, particularly for medical use, will be based on analysing complex patterns of neural activity. That is a task for which AI would seem well suited and which could replace much of the controversial animal experimentation that has taken place so far. Neuralink was reportedly placed under investigation by US federal authorities for animal-welfare violations late last year.
But on another, there is the issue of who could be given non-medical access to the other capabilities an AI/BCI combination might have. Would the differentiator be price or perhaps the selective largesse of those who control the IP? Cui bono?
It sounds familiar: new technology advancing towards exponential development and innovation, promising to do great good for many of the most disadvantaged, vulnerable and deserving – then comes the ‘but’.
The ‘Six Million Dollar Man’ famously portrayed superhuman speed through cheesy slow motion. As the possibility of a Six Billion Dollar Human moves closer, things are moving faster than many would like. Yet for those who genuinely could benefit from BCIs, things cannot happen quickly enough, and the case against that is very hard to make.
For now, existing regulation looks like a capable restraint and a way to distribute benefits soon in a generally acceptable way until wider issues are properly confronted and resolved. But – and here’s another familiar question – can it keep up?
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.
