A newly founded artificial intelligence lab, called Vicarious, wants to build the world’s first, unified artificial intelligence that can match human intelligence. This is not the first time we’ve heard companies or universities trumpet such ambitious goals, but considering who’s backing the project I can only entertain the possibility Vicarious might just do it. Entrepreneurs with great vision and a track record of backing successful companies have all hopped aboard, like Elon Musk (SpaceX, Tesla), Mark Zuckerberg (Facebook), Peter Thiel (Paypal, founder of venture capital and hedge funds worth billions), Jerry Yang (Yahoo! founder), Jeff Bezos (Amazon founder) and more.
Today, there are some pretty amazing AIs. For instance, Deep Blue and Watson – both from IBM – are some of the famous, garnering headlines when they beat the best human players of the time at chess and Jeopardy!, respectively. Then, there’s the deep learning machines that use neural networks to augment human behavior, like the trippy Google AI used to process images. The problem with these machines is that they’re too narrowed focused. Put them out of their domain, and they’re silly. Just a couple of days ago I wrote about how one of the best AI’s in the world scored no better than a four-year old on an IQ test – and that’s still remarkable.
“In contrast, Vicarious is building a single, unified system that will eventually be generally intelligent like a human. That means a machine that will be able to make sense of the world around it, building a complex and nuanced model of reality based on past experience and current sensory data,” said founder and computer scientist Scott Phoenix.
The company was founded in 2010, and in 2014 was awarded $40 million in funding, on top of the $70 million raised before. Since this isn’t a commercial venture – not yet, at least – Phoenix says this money is essential to keeping them afloat for at least a decade’s worth of research. That also means that they better come up with something to show if they want the funding extended for another decade. According to a report that surveyed 550 AI researchers, most believe there’s 50% chance a human-level AI might happen before 2050, and 90% by 2075. But how do does a machine get to the level of a human?
“The first prerequisite is processing power. A human brain, for example, has about a thousand times as many neurons as a frog brain. Whereas it took evolution about 250 million years to achieve a thousand time increase in processing power, our computers improve a thousand times every 10 years or so. Even today, we have a tremendous amount of underutilized computational power.
The more challenging task is understanding and replicating the function of the neocortex, the part of the brain that allows humans to learn and reason. Vicarious is building a mathematical model of the human brain that enables our systems to learn how to solve problems the way a person would,” Pheonix says.
The main challenge however lies in holes in our fundamental understanding of how the human brain works, something that’s out of Vicarious’ hands. So, the company depends on research elsewhere which might better explain how synapses join together to give rise to thought, or how a given atomic arrangement (matter) gives rise to self-awareness. So far, the best thing Vicarious has to show is a ReCaptcha solver – the kind used by websites all over the internet to filter out robots from humans. There’s more to humans than interpreting annoying faded text in boxes, though. Vicarious has its work cut out, but this is definitely something worth following. They seem very motivated.
Enjoyed this article? Join 40,000+ subscribers to the ZME Science newsletter. Subscribe now!