Cogitai has released Continua, a new software-as-a-service (SaaS) artificial intelligence platform that may herald a new opportunity for businesses to create self-learning AI systems that require far less training and human involvement than systems do today.
“We’re focused on an entirely different kind of AI that up to this point hasn’t been available commercially,” Mark Ring, CEO of Cogitai, told eWEEK. He says that most commercial AI applications today involve supervised learning where there is a human required to label a data set and the algorithm finds patterns in the data.
“It’s very powerful and has infiltrated the industry in many sectors,” said Ring.
But AI applications today are also limited because they don’t have the ability to make decisions and process what happens as a result of those decisions in a sophisticated way. Cogitai is focused on the “second generation of AI” where AI can make decisions on its own based on a reinforcement learning.
As an example, Ring said reinforcement learning is like what happens when you play a board game—a player makes a move, the opponent makes a counter move and the first player makes a move based on that new information. At the end of the game, you can see what the good moves and bad moves were and you—hopefully—learn how to play better for next time.
A New Generation of Self-Learning Applications
The Continua platform is now available for licensing (including a free, 40-day trial), and Cogitai said it’s already in use by a handful of large companies developing new AI-driven applications. These self-learning applications include “smarter” vehicles, web marketing, building management, AI software testing bots that learn to find and fix issues, on-board calibration and real-time energy management for semiconductors, and even AI-powered virtual coaches that offer personalized wellness and fitness advice.
The basic value proposition is that Continua can automate almost any repetitive decision process and only gets more effective as the number of decisions increases. The results can also be applied to similar use cases.
One of the top five car makers is using Continua’s platform to understand the way a car reacts to different environments and automatically adjust for greater fuel efficiency and performance. “For example, it will change the smart plug firing sequence in real time. We’re like an agent that lives in the car,” Dennis Crespo, vice president of marketing and business development at Cogitai, told eWEEK. “It doesn’t make it an autonomous vehicle, but it’s applying AI in a way that gives immediate, near-term benefits.”
Crespo says the same processes can be applied to robotics to pick a path in a warehouse, learn to navigate and pick up items as the robot develops its own skills that today typically require human training. Continua also sees opportunities in finance for more effective fraud detection where the system learns from past breaks and identifies new ones as they happen as well as the best actions to take for remediation.
Sony is a major investor in Cogitai. In addition to Ring, a pioneer in reinforcement learning, the executive team includes Vice President of Engineering Peter Wurman, who was the technical co-founder of Kiva Systems, known for its early use of mobile robotics in warehouses and distribution facilities. Kiva was acquired by Amazon.
"Since our investment in Cogitai in 2016, we have seen the team make tremendous progress in bringing their unique technology to market," Toshimoto Mitomo, senior general manager of the Startup Acceleration Division at Sony, said in a statement. "We believe self-learning AI as a SaaS platform will be a game changer and we will see many new opportunities to implement AI in a wide range of innovative business applications.”