While Generative AI is all the rage right now, what OpenAI, Microsoft and Google are doing may only be part of the story. There’s also the process of harnessing biology: the idea of using stem cells to create biocomputers that could be smarter and more energy-efficient than what we’re used to today.
Australian startup Cortical Labs popped up on the radar after Amazon CTO Werner Vogels recently flew to Australia to visit their lab, and even wrote about it as “exciting.”
It combines cortical synthetic biology and human neurons to develop what it claims is a class of AI known as ‘Organoid Intelligence (OI)’.
It has now raised $10 million in funding led by Horizons Ventures, along with LifeX (Life Extension) Ventures (a startup it covered last year), Blackbird Ventures, Radar Ventures and In-Q-Tel (the venture arm of the sector) CIA).
The company says it is already fulfilling orders for the technology.
How it works is by using clusters of neurons spun from human stem cells to create what it calls “DishBrain,” which is then bound in hard silicon and described as a biological intelligence operating system (BIOS).
Some observers say that this is the future of AI, because human neurons can be better than any digital AI model for general intelligence, because they are self-organizing and require much less energy consumption.
“The possibilities for hybrid AI to meet the artificial biology model are limitless, accelerating the possibilities of digital AI in a more powerful and sustainable way,” said Hon Weng Chong, founder of Cortical Labs, in a statement.
“Ultimately, being able to use these systems to better understand how neurons exhibit intelligence will open up many applications, including revolutionizing personalized medicine and disease diagnosis,” said Jonathan Tam of Horizon Ventures.
Cortical Labs’ technology first appeared in the October 2022 issue of the science journal Neuron, showing that neurons in a petri dish can be stimulated to play the computer game Pong.
This sounds simple, but as Weng Chong told me in an email, this will allow for the development and testing of new drugs and treatments, and “if you take your blood and put it into neurons, this drug discovery will be more personalized – the results will be customized just for you.”
In addition, competition for the space is low: “It’s not competing directly with anything because this is the first of its kind in the field of organoid intelligence. Organoid Intelligence has the potential to learn faster and use less power than any other AI system. GPT is so smart because it eats up the entire internet, so you or I don’t have to have good conversational skills.
“It took at least 10 years for us to get to where we are now after Geoff Hinton and Alex Krzyzewski made GPUs to do deep learning. We are still in the early days of this technology,” he added.
In the near future, he says, an immediate application is to efficiently drip a new drug onto cells – if the cells can’t play Pong anymore, they know the drug won’t work: “Not only the effectiveness, but better decision-making, but cognitive side effects (brain fog) can be explained because they’re now playing a game of Pong with neurons.” We can assess the ability to recognize the shape of cells.
He said the technology could be used to study dementia and even ‘brute force’ test the compounds we’ve discovered using quantum computing and generative AI.
In the future, “if the number and complexity of these neurons were measured, the results would be as familiar to us as whole-bodied organisms such as cats, dogs, or humans.”
Hold on to your hats, people.
