Instead of telling their company’s artificial intelligence design how to learn, Igal Raichelgauz and his team took a novel approach—they let a brain tell them how to build an AI.
A rat brain, to be specific.
CEO Raichelgauz and his fellow Cortica cofounders decided that the advanced, heavy computer technologies being used to recognize and quantify images and sounds were not only inefficient, but downright bad at what they were trying to do.
Those techniques “still fail to get to the level of a biological organism or even a baby when it comes to the ability to recognize objects and their different variations,” according to the Israeli entrepreneur.
While the human—or animal—brain has been an obvious inspiration for AI developers since the first notions of computer intelligence, the Cortica team took a different approach with that influence right from the company’s inception.
Working with a still-living piece of rat brain, the crew had the gray matter on a dish that gave them an electrical interface with nearly every neuron of the tissue.
This neuron-level input/output arrangement allowed the entrepreneurs to read responses from their electrical inputs to single, or multiple neurons, and various combinations over time.
“We started to play with it more like hackers, like engineers, trying to understand what it does before even answering the question of how it does that,” Raichelgauz says.
He and his early Cortica lab were reverse-engineering the brain.
They found that with some adaptation, a network of neurons could create what Cortica calls a conceptual signature—and could do so without supervision or training.
“Imagine if you have two images of the same person,” the CEO says. “The network will produce a sequence of numbers for each of the images and there will be a very high overlap between those two numbers, while if the images contained completely different objects the overlap would be zero, or minimal.”
So instead of the labor-intensive method of feeding images to an AI model and telling it “these are cats,” a new computational framework emerged where various cats were mostly similar but unique linear representations.
“It is not dependent on what you’re interested in doing with it,” Raichelgauz says. “Later you can associate it with what we want to label it and so on, but it has a very unique internal representation which does not depend on the task or the object of interest.”
In this case, the neural networks themselves are the building blocks for representation, not systems to be trained with our preconceived and inflexible definitions.
Equipped with the ability to create its own representations, this form of AI can learn more like a baby or child through context. But instead of from a crib, this young learner’s stimuli all come from the vast existing and updated knowledge on the web.
And while leaving a toddler alone would be considered bad parenting, the creators at Cortica see great advantages to their brainchild learning unsupervised. Crawling and exploring around a practically unlimited set of images and associated metadata, the technology needs no human in the loop for its learning.
“It’s not about trying to classify [images],” Raichelgauz says. “It’s more about visual discovery of the world and clustering and understanding what is what and then creating this association with the language.”
The outcome, a computational system his team designed, creates surprisingly accurate results, according to the cofounder. It deals with a couple billion potential numbers in a sequence, with any given image selecting only a hundred or two on average.
However, developing impressive artificial intelligence alone isn’t enough for the company.
Now that their AI has grown up a bit, it’s time for the Cortica team’s next step: applying that framework in a way that is helpful to potential customers.
The company has largely focused on image interpretation because the visual sphere is key in most engagements today for consumers.
But the group has also fed their AI other information like sounds, financial data and industry-specific medical imagery. The firm is in the process of looking for the right marketing strategies to launch its first app, and Raichelgauz says that’s on the very near horizon.
“When it comes to images, we want to create an experience using technology that’s really different from search or browsing or things that were designed originally for web or text,” he says. “I think we’re going to bring something very unique there.”
Raichelgauz won’t mention specific companies, but says Cortica is developing B2B partnerships in the mobility space and hints at the clear promise of the tech’s ability to recognize non-predefined patterns in tandem with autonomous vehicles.
It’s all a big leap from the company’s beginnings, but as the Tel-Aviv native says, Cortica “evolved slowly but confidently.”
With a background in electrical engineering, Raichelgauz first became interested in the brain viewing it from an electrical circuitry perspective.
Having seen the limitations of what were then some of the most advanced AI technologies in a previous job, he partnered with neuroscientist/biologist Karina Odinaev and vision and image scientist Yehoshua Zeevi to start Cortica.
They have since grown to 100 employees with offices in Israel, New York and Beijing.
“It’s really exciting to see how it works,” Raichelgauz says. “The problems we’re dealing with are really challenging.”
The likes of his company’s competition—Microsoft, Google, Amazon and other big players—have many more resources at their disposal.
“I think it’s pretty exciting to see even with those unfair conditions, we can achieve not just better results… but really fundamentally different results when it comes to context,” the Cortica co-founder says.
That underdog status, the unequal situations, feed the motivation of Raichelgauz.
As does converting the technology to the actual hands-on user experience—a milestone the company is starting to realize.
Reverse engineering the brain is no modest claim, and it took gray matter of all sorts to start Cortica on this path.
But while the intelligence of its product might be artificial, for Cortica the tangible results from years of hard work are finally becoming real.