December 2, 2023

While generative AI is hot right now, what OpenAI, Microsoft, and Google are doing may only be part of the story. And the process of using biology: the idea of ​​using stem cells to create biological computers that could be smarter and more energy efficient than the ones we use today.

Australian startups cortical laboratory It popped up on the radar after Amazon CTO Werner Vogels recently flew to Australia to visit their lab, and he even wrote an article about it, calling it “Intriguing“.

Cortical combines synthetic biology and human neurons to develop what it claims is a class of artificial intelligence called “organoid intelligence (OI).”

It has now raised a $10 million round led by Horizons Ventures, LifeX (Life Extension) Ventures (we covered its launch last year), Blackbird Ventures, Radar Ventures and In-Q-Tel (CIA’s risk division).

The company said it was already fulfilling orders for its technology.

How it works is that it uses clusters of lab-grown neurons from human stem cells to form what it calls a “DishBrain,” which it then attaches to hard silicon to create what it describes as a biologically intelligent operating system (biOS).

Some observers say this is the future of artificial intelligence, as human neurons may be better than any digital AI model of general intelligence because they program themselves and require far less energy expenditure.

Hon Weng Chong, CEO and founder of Cortical Labs, said in a statement: “The combination of hybrid AI and synthetic biology models can unlock endless possibilities and accelerate the advancement of digital AI in a more powerful and sustainable way. possibility.”

Horizons Ventures’ Jonathan Tam said: “Ultimately, by being able to use these systems to better understand and ultimately harness how neurons display intelligence, it will open up a host of applications, including revolutions in personalized medicine and disease detection. “

Cortical Labs’ technology, which first appeared in the scientific journal Neuron in October 2022, demonstrates that neurons in a petri dish can be encouraged to play the computer game Pong.

It sounds trivial, but as Weng Chong told me in an e-mail, it could help in the development and testing of new drugs and therapies, and, furthermore, “If you make neurons from your blood, the drug’s Discovery will become more personal—the results will be tailored just for you.”

He also says competition in the field is low: “It’s not going to compete directly with anything because it’s a first for organoid intelligence. Organoid intelligence has the potential to learn better than any other AI system out there.” It’s fast and consumes a lot less energy. GPT is very smart because it sucks up all the internet, but you or I don’t have to have pretty good conversational skills for us.”

“It’s taken us at least 10 years since Geoff Hinton and Alex Krizhevsky used GPUs for deep learning. We’re still in the early days of the technology,” he added.

In the near term, he says, one immediate application would be to test a new drug effectively by dropping it on cells—if the cell can no longer play ping-pong, you know the drug isn’t working: Well identified, but also able to shed light on the cognitive side effects (brain fog) because we now have a potential method of cognitive profiling in the form of neurons playing a game of ping-pong.”

The technique could also be used to study dementia, he says, and even “brute-force” test compounds we discover using quantum computing and generative artificial intelligence. “

And possibly in the farther future “if the number and complexity of these neurons are scaled, the end result will be the fully embodied organisms we are familiar with, such as cats, dogs or humans.”

Hold on to your hats, people.