Artificial intelligence might now be solving advanced mathematics, performing complex reasoning, and even using personal computers, but today’s algorithms could still learn a thing or two from microscopic worms.
Liquid AIA startup out of MIT, will today reveal several new AI models based on a new type of “liquid” neural network that has the potential to be more efficient, consume less energy and be more transparent than those that underpin everything from chatbots to image generators and facial recognition systems.
Liquid AI’s new models include one to detect fraud in financial transactions, another to control autonomous vehicles, and a third to analyze genetic data. The company promoted the new models, which it is licensing to outside companies, at an event today at MIT. The company has received funding from investors including Samsung and Shopify, both of which are also testing its technology.
“We are expanding,” he says Ramin Hasanico-founder and CEO of Liquid AI, who co-invented liquid networks when he was a graduate student at MIT. Hasani’s research was inspired by the C. elegansa millimeter-long worm usually found in soil or decaying vegetation. The worm is one of the few creatures whose nervous system has been completely mapped and is capable of remarkably complex behavior despite having only a few hundred neurons. “It was once just a scientific project, but this technology is fully commercialized and ready to bring value to companies,” says Hasani.
Within a normal neural network, the properties of each simulated neuron are defined by a static value or “weight” that affects its activation. Inside a liquid neural networkEach neuron’s behavior is governed by an equation that predicts its behavior over time, and the network solves a cascade of linked equations as the network operates. The design makes the network more efficient and flexible, allowing it to learn even after training, unlike a conventional neural network. Liquid neural networks are also open to inspection in a way that existing models are not, because their behavior can essentially be rewinded to see how it produced a result.
In 2020, researchers showed that such a network with just 19 neurons and 253 synapses, which is remarkably small by modern standards, could control a simulated self-driving car. While a normal neural network can analyze visual data only at static intervals, the liquid network captures the way visual information changes over time very efficiently. In 2022, the founders of Liquid AI discovered a shortcut That made the mathematical work needed for liquid neural networks feasible for practical use.