Creator of a brain-computer interface, Nextmind is working to industrialize his invention by preparing a new fundraiser. Its founder tells the Minority Report all about his project.
JDN. Nextmind created the first portable device to control an interface by thinking. How did this brain-computer interface come about?
Sid Kouider. Nextmind was created in 2017 in the laboratory of the Normal Sup school. The work mainly focused on neurophysiology and brain wave analysis. This work aimed to understand brain mechanisms such as attention, perception, etc. Recent advances in Machine Learning and sensor miniaturization have enabled us to design a first prototype. Our neural interface makes it possible to perform simple actions on a computer screen, such as changing channels, moving objects or lowering the volume, only thanks to his brain. Nextmind is now a Parisian start-up with around twenty employees.
How does this technology work, which has the particularity of being entirely based on the analysis of brain activity without resorting to Eye Tracking technology?
The idea is to decode, in real time, the elements on which a user is focusing. In other words translate the cerebral signals coming from the visual cortex. Nextmind has developed its own hardware device because EEG headsets (electroencephalogram, editor’s note) that one finds in the hospital environment are often quite bulky, and require gel as well as a rather long installation time. Our device, equipped with sensors and connected via Bluetooth, can be installed in a few seconds with a band around the head. It is non-invasive and measures the electrical activity of the brain when a person focuses on a specific element. We will first capture the brain waves to decode them. Our machine learning algorithms will then analyze this data in order to convert it in real time into action on the computer.
At CES20, you announced the sale of 1000 first development kits for $ 399. Where are you today and what is the profile of buyers?
“We are targeting the video game, virtual reality, disability market and the Internet of Things sectors”
These kits are now sold in Europe and the US. We are also seeing growing interest from the United States. Regarding the profile of buyers, for the time being, these are mainly developers from the video game and virtual reality sector. Rather creative profiles, who seek to create immersive experiences. We work for example with VR artists but also with people who seek to create new experiences in the field of cinema or events. At the start of the school year, three artistic projects should be created with our technology. These projects are useful for demonstrating our technology and thus allowing the general public to test our neural interface.
Why this focus on the video game sector?
Video games have always been conducive to innovation. This has been the case in the past with the integration of the Kinect’s gesture recognition technology, for example. It is a market of early adopters with a real appetite for technology and for the discovery of new experiences.
Besides video games, what other use cases do you think of?
We discuss and do co-development with players in augmented reality and virtual reality. We are also considering adapting our technology to the disability market. Another interesting use case is IoT. Instead of interacting only with a screen, a user could interact, via his brain, with connected objects. Imagine, for example, that you could open doors, turn on the light, turn on your tap or validate your ticket on the bus just by thinking. This is why we are working a lot on the miniaturization of our device. Some car manufacturers are already considering integrating it into the autonomous car of tomorrow. Thus, not only would the passenger not put their hands on the steering wheel, but they would no longer put their hands on the interface either.
What revenue model do you envision for commercializing and distributing this technology?
“Car manufacturers are considering integrating this technology into their future autonomous cars”
Our model is a licensing model. It’s pretty close to that of Primesense, the Israeli company that developed the Kinect’s motion-sensing technology. The company had licensed its technology to Microsoft while developing it for other uses. We want our technology to be integrated into the technology roadmap of as many companies as possible, whether through our own device or in third-party devices. The goal is that our neural interface can reach the widest audience. Having more users will also allow us to improve the performance of our machine learning algorithms, which are fed with this completely anonymized data.
Will this technology be able to read minds in the future? Could we, for example, imagine no longer needing to use a keyboard to write text?
At the moment we are not yet decoding thoughts. We only translate a person’s active intentional focus into action. We have already taken an important first step by allowing a computer to be controlled with its brain in real time without an invasive device. The second step will be to decode the imagination. This is a big problem from a neuroscience and machine learning perspective, but one that we should be able to solve over the next few years.
We understand better and better how the brain works, and machine learning now offers us new possibilities. However, algorithms and sensors are not yet sufficiently evolved. The brakes are therefore more technological than theoretical. But it is, in my opinion, a matter of time. Ethics-related questions will then arise that must also be taken into account.
After raising nearly 4.6 million euros in December 2018, is a new fundraising planned? Nextmind must already generate interest from potential buyers. Is your desire to remain independent?
We have indeed started discussions for a new fundraising. We’ll probably be announcing something in the coming months. These funds should allow us to scale our activity. The objective will be to continue to improve the technology and to recruit. Regarding a possible takeover, we do not think about it at this stage. We are focused on developing the next version of our product. Our goal is to democratize neurotechnologies by creating the best platform in the world in this sector. For Nextmind, it is a question of continuing to innovate.
Sid Kouider is a French neuroscientist and CEO of Nextmind. Between 2008 and 2018 he directed the Lab on “Consciousness and the Brain” within the Ecole Normale Supérieure. Among the various works carried out within this lab, the Nextmind project will be born in 2017, a brain-computer interface whose technology makes it possible to decode the electrical activity of the brain to convert them into an action on the computer. He holds a Phd in Cognitive Sciences obtained from the Ecole Normale Supérieure