NewsBite

Breakthrough: AI can now be trained on human thoughts thanks to Synchron/Nvidia tie-up

A company spun out of Melbourne University has partnered with Nvidia and Apple to build an AI model which can help those with paralysis regain function.

Rodney Gorham from Moonee Ponds in Melbourne is among the first to use Synchron's Chiral — an AI model which can be directly trained by human brains.
Rodney Gorham from Moonee Ponds in Melbourne is among the first to use Synchron's Chiral — an AI model which can be directly trained by human brains.

Synchron — a company spun out of Melbourne University — has partnered with Nvidia and Apple to build an artificial intelligence model which can be directly trained by human brains.

The model — called Chiral, Greek for hand — is designed to help people with paralysis regain movement and perform everyday tasks, albeit in the digital realm.

Nvidia senior director of business development David Niewolny said: “We are witnessing a new computing paradigm, where technology is no longer just responding to people — it’s empowering them in entirely new ways”.

Former IBM salesman Rodney Gorham, 63, from Moonee Ponds in Melbourne, is among the first people to use the new technology from Synchron, which counts Bill Gates and Jeff Bezos among its ­investors.

How Synchron's brain computer interface works.
How Synchron's brain computer interface works.

Mr Gorham has amyotrophic lateral sclerosis, or ALS, a progressive neurodegenerative disease which affects nerve cells in the brain and spinal cord, leading to the loss of muscle control and ­paralysis.

Mr Gorham cannot speak or use his hands, but thanks to Synchron’s technology he can still perform everyday tasks such as feeding his dog, playing music, turning lights on or off or directing his robot vacuum to clean the floor — using just his thoughts.

Chiral works by connecting Synchron’s brain computer interface (BCI) with Nvidia’s Holoscan platform — an AI computing platform for developers to build high-performance streaming AI apps for medical devices which can be displayed on Apple’s Vision Pro spatial computer.

Synchron co-founder and chief executive Tom Oxley.
Synchron co-founder and chief executive Tom Oxley.

The system works by using an implanted device to read the electrical activity in the brain which controls movement. It uses Apple Vision Pro to “see” the signals and then converts them using the Nvidia computer into instant commands from the Apple Pro for digital devices. Synchron co-founder and chief executive Tom Oxley said the technology effectively allowed people with paralysis to interact with the physical world simply by thinking.

“We are building a brain foundation model using generative pre-training techniques that learn directly from neural data — abstracting human cognition at its source — to create features that improve our user’s lives,“ Dr Oxley said. “This is possible because of our ability to scale large datasets, by making BCI as common as a stent insertion.”

Synchron has developed a more advanced and less-invasive system than Elon Musk’s Neuralink, which has raised more than $600m to invest in research, but suffered early setbacks which almost derailed its first patient’s implant after its performance deteriorated.

Synchron has integrated its BCI in Amazon Alexa — the tech titan’s voice assistant — allowing people with paralysis to control their smart home devices such as lights, speakers and TVs.

Chiral, over time, will learn Mr Gorham’s brain signals and become faster at responding and recognising his intent, allowing him to control the environment in real time. For example, when the system sees his pet dog, it asks if it needs to be fed.

Originally published as Breakthrough: AI can now be trained on human thoughts thanks to Synchron/Nvidia tie-up

Add your comment to this story

To join the conversation, please Don't have an account? Register

Join the conversation, you are commenting as Logout

Original URL: https://www.dailytelegraph.com.au/business/breakthrough-ai-can-now-be-trained-on-human-thoughts-thanks-to-synchronnvidia-tieup/news-story/a5b548019a086ad74e742c4da9110400