Synchron Is Integrating ChatGPT into Its BCI System for Disabled
TapTechNews August 6th news, to help disabled people and those with motor impairments use mobile phones, computers and other devices more easily, the American brain-computer interface (BCI) company Synchron is attempting to integrate ChatGPT into its brain-computer system, which the company claims is a global first.
The implant of Synchron, known as the stentrode, is inserted into a blood vessel near the brain's motor cortex. The price of this implant is between 50,000 to 100,000 US dollars (TapTechNews note: currently about 357,000 to 714,000 RMB), and it does not require a craniotomy, which is comparable to the cost of other implanted medical treatments such as pacemakers or cochlear implants.
Mark is one of only 10 people in the world who have received an implant of Synchron's brain-computer device. He was diagnosed with amyotrophic lateral sclerosis (ALS, commonly known as Lou Gehrig's disease) in 2021, and he said his hand is almost completely unusable at present.
It is said that although ordinary BCI devices can also help him operate the cursor for typing and clicking operations, it is very, very slow. But now, he can have rapid conversations with the help of AI prediction. The AI occasionally says bad words, although I do that occasionally too, he said with a smile.
Synchron's CEO Tom Oxley said that the company has experimented with various different AI models in the past year, but the release of OpenAI's ChatGPT-4o in May 2023 offers some interesting new possibilities.
The o in ChatGPT-4o stands for omni, indicating that this version can simultaneously receive text, audio and visual inputs to generate output.
In the OpenAI demonstration, what attracted Oxley the most was that it showed a scene of a visually impaired person finding their way in the city, asking the AI to describe his surroundings and even helping him call a taxi.
Oxley believes that a similar scenario may also exist in the future of BCI: allowing large language models like ChatGPT to receive relevant context in the form of text, audio and visual to provide relevant prompts that users can select through BCI. Oxley also said that the company will not be bound to any specific large language model.