7 Ways Brain-Computer Interface Companies Are Using AI
While we’re still a couple of months out from the release of the final installment of the Star Wars triple trilogy, you can already buy tickets to watch the good guys blow up the Death Star for the fifth or sixth time (we’ve lost count). The 42-year-old franchise has raked in more than $9.3 billion in box office sales alone, begging the question: Why is it so popular given that George Lucas couldn’t direct his way out of an escape room and the plot is older than Yoda’s underwear? Simple: the Jedi. Who doesn’t fantasize about being able to control the Force with your mind, especially after smoking some Jedi Kush? That’s our not-so-subtle introduction to the theme of brain-computer interface (BCI), an emerging technology that melds the human mind and machines to help people walk again, play the latest virtual reality game, or turn you into a supergenius.
CTRL-Labs Goes Off the Market
BCI tech, also referred to as brain-machine interface (BMI), has been back in the news lately, largely thanks to the acquisition of a New Yawk-based startup called CTRL-Labs by Facebook, a deal that Bloomberg estimated is worth between $500 million and $1 billion. That’s quite the premium, given that the four-year-old company had raised $67 million in disclosed funding from investors like Google and Lux Capital. Of course, Facebook isn’t afraid to spend loads of cash on an acquisition. Five years ago, it doled out $2 billion for VR startup Oculus. The purchase of CTRL-Labs is expected to augment the social media giant’s own efforts in brain-computer interface tech development, particularly with an eye toward VR and related applications.
We devoted a whole story to CTRL-Labs last year, detailing its machine learning platform that digitizes electrical activity in the user’s arm in order to control a computer avatar. In other words, your hand becomes the mouse or keypad, using only minimal gestures, like a symphony conductor directing an orchestra with a flick of the wrist. Over time, the algorithms get better and better at reading and responding to the firing neurons.
Neuralink Proves It’s Not Just Monkeying Around
CTRL-Labs is just one example of a company using artificial intelligence to improve the interface between man and machine. Another BCI company back in the news recently, Neuralink, is not just using AI machine learning to link brains and machines. The Elon Musk-backed venture wants to give humans the computing power of AI itself so that Homo sapiens don’t meet a Terminator-esque ending. Toward that goal, the company has raised $158 million, most of it coming from Musk himself. This summer Neuralink announced it successfully tested the technology, which involves stitching slender electrodes directly into the brain, on monkeys. Musk said Neuralink will move up the evolutionary chain to humans by next year, with hopes to treat brain and spinal cord injuries before creating a race of
unholy cyborgs enhanced humans.
The Future Machines of War
The shadowy government agency simply known as DARPA has been funding research into BCI for going on two decades. More recently, it has been pushing programs that leverage machine learning and other AI technologies to adapt BCI tech for future military purposes.
For example, earlier this year, DARPA made six awards under its Next-Generation Nonsurgical Neurotechnology initiative, which seeks to:
Develop high-resolution, bidirectional brain-machine interfaces for use by able-bodied service members. These wearable interfaces could ultimately enable diverse national security applications such as control of active cyber defense systems and swarms of unmanned aerial vehicles, or teaming with computer systems to multitask during complex missions.
While most of the projects are being run out of university labs at prestigious institutions like Carnegie Mellon and John Hopkins, one of the awards went to a nonprofit company (is that really a thing?) called Battelle. The contract, worth up to $20 million, would build upon the company’s NeuroLife technology that uses machine learning to help paralyzed patients regain conscious control of their fingers, hand, and wrist. The new project is called Brain System to Transmit Or Receive Magnetoelectric Signals (BrainSTORMS) and uses a nano-transducer that is injected into the body before finding its way to a specific area of the brain. The nano-transducer communicates with a helmet-based transceiver using magnetoelectric nanoparticles, converting electrical signals into magnetic signals. So, yes, it sounds like Battelle is creating a hybrid Magneto-Professor X superhero-villain. The original NeuroLife was able to help a man with spinal cord injuries move his hand again.
In a separate $1 million award from DARPA to Battelle, the company is working on improving the algorithms that help decode and translate brain activity into action. That money is coming from a $2 billion DARPA program called the “AI Next” campaign. As in the next thing you know, an autonomous machine has just blown you to bits.
Brain-Computer Interface Startups
Let’s return to peacetime and profile a few new BCI startups that we haven’t covered before to see where this emerging market, expected to be worth about $1.4 billion by next year, is headed.
Founded in 2011, BrainQ, based in the Holy City of Jerusalem, has raised about $8.9 million, largely recruiting its AI team from the elite intelligence units of Israel’s security forces. Its AI algorithms identify high-resolution patterns in a patient’s brain waves, which are interpreted and translated into a personalized electromagnetic treatment for people suffering from disabilities that occur as a result of stroke or spinal cord injury. BrainQ claims its algorithms have been trained on one of the largest known databases for motor tasks, and the technology has already shown promise in animal experiments and some early human clinical trials.
Founded in 2017, Paris-based NextMind has raised $4.6 million in disclosed funding. Among the investors is a guy named Alstrup Johansen, who sold his eye-tracking tech company to Oculus/Facebook in 2016. NextMind is attempting to develop brain-sensing technology for virtual and augmented reality headsets, similar to a startup we covered previously called Neurable. The noninvasive tech device would allow users to play games and control computer devices in real-time using just their thoughts thanks to machine-learning algorithms that translate brain waves direct digital commands, presumably in a similar fashion to CTRL-Labs.
Founded in 2014, Santa Barbara, California-based Cognixion raised a $1 million Seed round last year to develop an AI-powered BCI interface that helps nonverbal people communicate using their mind by translating brain waves into words. As a step toward that goal, the company released an app earlier this year that according to one report “uses eye-tracking technology to allow users with limited mobility to navigate the screen with their eyes and select letters or words by blinking or holding their gaze. It also leverages facial recognition software to pick up on smiles and other muscle movements.” That puts it in competition with companies like Tobii (TOBII), a Swedish firm that derives about two-thirds of its revenue from assistive communication applications using eye-tracking tech. The pro version of Congixion’s Speakprose app costs $39.99 per month but only works with Apple mobile products.
Founded in 2011, San Francisco-based Emotiv has reportedly raised about $1.6 million. A previous incarnation of the company actually existed back in 2003 in Australia, where founder Tan Le and her family ended up as refugees after leaving their native Vietnam. Like other BCI companies using AI, Emotiv has trained its machine learning algorithms to translate electrical patterns – in this case, using a headset – into commands. The company gained a certain amount of fame in 2017 when a quadriplegic man used its technology to drive a Formula One race via an on-board computer that turned his thoughts into a need for speed. Emotiv sells a number of hardware and software products, including its flagship EPOC headset and EmotivPRO software that target markets as diverse as brain research and consumer insights.
We always assumed that we’d be driving flying cars well before we could perform a few Jedi mind tricks, but recent advances in brain-computer interface technology using artificial intelligence suggests that we might be able to fly with just our minds – at least in the virtual world. Yet these and other neurotech companies and researchers are finding real success in connecting the human brain to machines for electroceutical treatments for all types of conditions, including sleep apnea and epilepsy. Even more exciting is the potential to help people walk or even run again.