Gabe Newell, the famous CEO of Valve, has been working for several years on a new brain-machine interface (BCI) through his startup Starfish Neurosciences. A little over a week ago, this entity finally emerged from the shadows to reveal some snippets of its roadmap, announcing the arrival of a first chip by the end of the year. Here's how it differs from other BCIs on the market, starting with Neuralink.
The Steam chief executive's interest in neuroscience is nothing new. As The Verge points out, he has been exploring this topic from different angles for years. More than ten years ago, he organized research on the physiological response of the human body to video games. We also remember GDC 2019, where the company publicly expressed its interest in brain-machine interfaces applied to games. But this hobby is not at all part of the objectives of his startup Starfish Neurosciences.
A project focused on medicine
This is a significant difference from industry leaders, such as Neuralink. Elon Musk's company, for example, sees this technology as a way to augment human capabilities, notably through the integration of different forms of artificial intelligence. It has already demonstrated this concept. Its first and third patients, suffering from diseases that significantly impact their mobility, are, for example, able to use their computers with their thoughts, whereas they were previously totally dependent.
Starfish, on the other hand, seems entirely focused on strictly medical and therapeutic applications rather than utilitarian ones. Its future products, for example, will aim to modulate brain activity with great precision to treat bipolar disorders, various forms of depression, Parkinson's disease, epilepsy, and chronic pain in the longer term. The company also mentioned a system capable of neutralizing tumors by "cooking" them with great precision.
A tiny, energy-efficient chip
The other big difference is that Starfish is focusing on minimizing the invasiveness of its future device. In its presentation post, the firm explains that it wants to reduce the size of the components so that the whole thing fits into a 2x4 millimeter envelope.
And it's not just the physical dimensions that will be reduced to the bare minimum. The same goes for power consumption; according to the company, it should be around 1.1 milliwatts, almost six times less than Neuralink's N1 chip. This will notably eliminate the dependence on the internal battery and replace it with an external wireless power system – a bit like those found in wireless smartphone chargers, but smaller.
A distributed architecture
The last major difference directly concerns the chip's architecture. Those from Neuralink, Synchron, and others all share one thing in common: they are designed to be implanted at a single site, which is then stimulated with great precision thanks to a very dense network of electrodes. Starfish, on the other hand, relies on a distributed approach. In practice, the concept is based on several distinct sub-interfaces implanted in different regions of the brain – a bit like the arms of a starfish, hence the name of the company. This will make it possible to work not at the level of a brain area, but of an entire neurological circuit. A potentially decisive advantage in the treatment of complex neurological disorders like Parkinson's.
For now, Starfish still has a long way to go to catch up with its competitors. Several companies, such as Neuralink and Synchron, to name a few, already have relatively mature chips that are currently undergoing clinical trials. The former has already implanted its device in at least three patients. The second is even more advanced: after a first conclusive trial on six patients, it is now setting up a larger-scale trial for its Stentrode chip.
Gabe Newell's firm, on the other hand, has not yet built its first chips. These should arrive by the end of the year according to the post mentioned above, but this is only a first step. Even if the initial tests are satisfactory, it will likely be a few more years before the first clinical trials on humans. It will nevertheless be interesting to follow the trajectory of this startup, whose strategy is quite different from that of the current big players.
0 Comments