This is a guest post written by entrepreneur, Singularity University alumnus and teaching fellow Rod Furlan. He is an independent researcher in the fields of artificial intelligence, quantitative finance and high-performance computing. Passionate about technology and a serial autodidact, Rod has been to many high-tech industries always in search of the next big challenge. He is currently working on funding solutions for risky and highly-disruptive R&D projects with the intent of accelerating the singularity timeline.
I have just returned from a X PRIZE Foundation workshop on brain-computer interfaces (BCI) at MIT. The workshop brought together over 50 leading experts, students and enthusiasts with the objective of brainstorming ideas for an X PRIZE competition to accelerate the development of BCI solutions. During the course of this fantastic two-day event we had the opportunity to explore the many possibilities and difficulties of designing and implementing devices capable of communicating directly with the human brain.
The X PRIZE Foundation is a nonprofit organization whose mission is to foster innovation through competition. On October 4, 2004, the X PRIZE Foundation captured the world's attention when they awarded the largest prize in history, the $10 million Ansari X PRIZE, to Scaled Composites for their craft SpaceShipOne - the first privately built spacecraft capable of carrying three people to 100 kilometers above the earth's surface, twice within two weeks. With that single flight, a new industry was born and Spaceflight was no longer the exclusive realm of government.
Five years later I now find myself at MIT discussing how we could engineer a similar competition to conquer not outer space, but the inner space of the human brain. We had the opportunity to hear from the field's leading experts and to explore the many possibilities for competitions that could foster accelerated development of technology that would allow direct brain-to-machine and brain-to-brain communication.
Our deep-future vision for the technology was crystal clear: we want to give vision to the blind, new bodies to disabled people and maybe a GPS sixth sense to the willing. We want to communicate with each other and with our technology using thoughts alone - the possibilities are endless.
At the X PRIZE BCI workshop however, the big questions were not around the possibilities for the technology but around which steps could we take now to accelerate the pace of progress. In a way, we were trying to engineer the future itself as we considered different challenge modalities and incentives.
As Peter Diamandis, Chairman of the X PRIZE Foundation, would say: you will get what you incentivize for. There were many discussions about what we should focus on. Should our primary goal be to create incentives for the development of medical devices that would improve the lives of people living with devastating conditions? Or should we place our focus on the opportunities for human enhancement? Or maybe somewhere in between?
More importantly, a prize is only a good thing if someone wins it. To guide the process towards a winnable challenge one important constraint was imposed: we should only entertain ideas that could be won within a decade.
To ignite the brainstorming process, during the first day of the workshop we listened to several keynote presentations. First Peter Diamandis shared with us a bit of the story and ethos of the X PRIZE (watch his TED talk). He was followed by Ray Kurzweil and his inspiring vision for the future of BCI technology (watch it). Then after a short lunch break we had the opportunity to hear from John Donoghue (founder of Cyberkinetics), Gerwin Schalk (BCI2000) and Ed Boyden from the MIT Synthetic Neurobiology Group.
After the keynote presentations we were divided into four breakout discussion groups to explore different aspects of BCI technology: Input/Output, Control, Sensory and Learning. A few hours into the afternoon, the Learning group dissipated as participants realized that given the proposed time frame of the BCI X PRIZE it would be unlikely that anyone would be able to develop solutions for augmented learning by any means other than stumbling into an unexpected breakthrough by sheer luck.
The main challenge we would need to overcome to enable augmented learning is that talking to the brain and purposefully altering the brain to implant a memory or a skill are two fundamentally different tasks and we currently lack significant understanding of how to manipulate neuronal topology in a useful way.
That afternoon all groups focused on forecasting the advances of BCI technology up to 40 years in the future and identifying the most important milestones along the way. Each milestone achievable within a decade from now being a potential candidate for an X PRIZE competition.
Early next morning we had two more presentations about the current cutting edge of BCI technology. First Dean Pomerleau from Intel Labs shared with us details about his latest research on thought recognition. He was followed by Christopher deCharms' presentation on real-time fMRI scanning (watch his TED talk).
After the presentations it was time to start pruning ideas from our lists. Each group then worked towards identifying and presenting their best 3 candidates for either a X PRIZE (achievable within 8-10 years) or a X CHALLENGE (achievable within 2-3 years).
While we still have significant technical and scientific hurdles ahead of us, given the current pace of progress it is reasonable to expect that robust, albeit limited, implanted BCI solutions will be widely available commercially within a 10 to 20 year time frame. My definition of widely available in this case would be a market penetration similar to what cochlear implants have today.
Invasive vs. Non-Invasive
As non-invasive interfaces are generally limited to reading brain states, it is unlikely they will be able to evolve into robust input and output solutions. Consensus among the experts in the room was that EEG is probably a dead end because while it provides great temporal resolution, its maximum achievable spatial resolution will probably fall short of the requirements of future applications.
One big factor that is currently slowing down progress of implanted, invasive devices is that while animal models are great to get research off the ground, without human subjects it would be impossible to develop powerful interfaces for human use. However current FDA regulations in the United States don't make it easy to experiment in humans. Basically, unless someone is already going through open brain surgery for some other reason it is unlikely researchers would be able to implant experimental BCI devices inside their skull.
I am personally fascinated by regulatory and moral asymmetries - consider the example of test pilots. They put their lives at risk to test unproven aircraft and governments are generally fine with it. There isn't a heated discussion about it being a big ethical dilemma, a few test pilots will die so we can develop new aircraft - and we accept that because want new aircraft to be developed. When it comes down to medicine however, it is apparently a big deal to risk someone's life to further medical research - even if the subject is willing to accept the risk.
Is geo-arbitrage the solution?
Fast Company recently published a very interesting article on how a stem cell startup used geo-arbitrage to sidestep the FDA's clinical testing requirements. As technological progress accelerates, the pressure to decrease the time-to-market of new implantable devices could make this practice widespread. Unless there are changes to the FDA approval process to accommodate shorter product cycles, researchers in the USA might soon find themselves dealing with a situation that is much like the game theory problem known as "the prisoner's dilemma": even though the best possible global outcome can only be achieved if everyone goes through the FDA approval process, individual markets would eventually be forced into overseas testing one-by-one after a few key players "defect" in order to gain a short term edge.
The alternative: making "intrusive" not so intrusive
FDA regulations exist mostly to mitigate risks to patients. New implantation techniques that significantly reduce risks could become a practical alternative to geo-arbitrage and medical tourism.
For example, a mobile camera robot was recently developed to provide the ability for a single incision biopsy procedure. Using sophisticated computer-aided navigation and motion compensation for the movement caused by heart beats, researchers were able to extract deep tissue from living organisms with minimized risk and minimal collateral damage.
This same technology could one day be used to implant electrodes directly into the brain through a very small hole in the skull or even through softer tissue inside the nose or around the eyes, thus greatly reducing the risks involved with the procedure.
What is next?
The MIT workshop marked the beginning of the development of the BCI X PRIZE. A typical prize development process takes 8-14 months to complete. In the meantime the foundation will be conducting interviews with experts, governments, potential competitors as it strives to mature the competition's goals and rules as well to pursue donors for the prize purse. Once the funding is secured and the requirements for the prize are finalized, the race will be on for companies and teams around the world to compete to make a major breakthrough in BCI.