“People are beginning to wake up to the fact that the industry, the experts, the media and the investors have overhyped artificial intelligence. It’s failing.”
If you are familiar with the world of technology, it is difficult not to hear about the “rise” of artificial intelligence, deep learning and machine learning, over and over. Industries from manufacturing to security to even agriculture and education are now incorporating or are expected to incorporate several aspects of these new technologies in their functioning. This, in turn, is expected to result in the growth of a new breed of devices and solutions that are termed “smart”.
But what if we are heading in the wrong direction? What if the whole concept of artificial intelligence as we know it today is not actually intelligent at all? Although these questions might be dismissed as exaggerated skepticism by many technology enthusiasts, some experts stress that these are in fact far more relevant than they seem.
“People are beginning to wake up to the fact that the industry, the experts, the media, and the investors have overhyped artificial intelligence. It’s failing.”
-Sean Everett, Founder and CEO, PROME
Enter “biologic intelligence”. A next-generation artificial intelligence (AI) that goes beyond the capabilities of deep learning. It requires no training data, no internet connection, no massive cloud computing, and certainly no data scientist’s time to annotate data and build one-off AI models. Biologic intelligence emulates an animal’s brain and nervous system into software and robotics. This new form of artificial general intelligence runs on top of deep learning and machine learning-based systems to understand and act upon changing and unlabeled data in real-time.
A California-based company, PROME, has been working on this for six years and claims to have successfully developed a biologic intelligence-based solution that is faster, more powerful and will start working the moment it’s turned on, regardless of the environment. Speaking to asmag.com, PROME's Founder and CEO, Sean Everett said that “it’s the first step towards the holy grail of true artificial general intelligence.”
“I’ve been studying math my entire life, which is all that the state-of-the-art artificial intelligence and deep learning is,” Everett said. “It’s not intelligent at all. It’s just an equation + human time. Timothy Busbice, the CTO and other founder [of PROME] has been staining and slicing brain tissue since his days in undergrad 30 years ago. Since then, he’s been developing artificial neurons as a fun hobby. But we both knew that if we could copy the intelligence system of humans and animals, we would have real artificial intelligence on our hands. Not that mathematic stuff. But rather, biologic. And if we could do that, we could apply it to every industry and push this human race forward at blinding speed. So, here we are.”
PROME is software only, but is environment and chip agnostic. A few thousand lines of highly performant code, presented as an executable file that can run on device, at the edge, on premise or in a cloud environment, depending on the customer’s needs.
The company is currently talking to global tier 1 customers about using their product to help solve problems across industries. These include detecting machinery issues in factories before they occur to prevent downtime, autonomous obstacle avoidance in transportation, predicting and preventing seizures that affect one percent of the human population in health care, and executing optimal trades and trading strategies in capital markets. Biologic intelligence can also function as the brains and nervous system of lifelike service bots for uses such as security, education, rehab, medical, home assistance, elderly care and factory automation. It can even replace human auditors in the accounting field, detecting anomalous transactions as they occur in real-time, according to Everett.
“People are beginning to wake up to the fact that the industry, the experts, the media, and the investors have overhyped artificial intelligence,” Everett added. “It’s failing. Google ‘limits of deep learning’ and you’ll see an emerging media trend. We exist beyond that place where the status quo hits a brick wall and fails. Instead of having humans manually do a bunch of stuff with data and equations, shouldn’t software be doing that instead? The magic is almost lost because 99 percent of the world bet on the wrong horse.”