Are bionic-people becoming a reality?  

When AI enhances human capabilities and provides lost life opportunities

Ido Kenan, journalist

[maxbutton id=”1″ ]

MobileEye is a driving assistant that uses AI to watch the road for what drivers might have missed, identify dangers and alert the drivers. Could that mechanism be improved to assist people in their lives? OrCam thinks so. In 2015, the company launched MyEye, a wearable camera with a speaker which, according to OrCam’s site, is “designed to assist people who are blind, visually impaired, or have a reading disabilities [sic]” by recognizing faces, currency and products and conveying that knowledge to its users, as well as reading aloud texts from varied surfaces. Unsurprisingly, OrCam and MobileEye share their co-founders, Prof. Amnon Shashua and Ziv Aviram.

OCR (optical character recognition) existed before, but it was cumbersome and took time to process. Dr. Yonatan Wexler, OrCam’s Executive VP R&D, says AI has made it possible for MyEye, which is about the size of a finger, to read texts aloud in real time. “We have made the act of reading a technicality,” he says, disagreeing with critics who claim that machines reading to us are ruining reading for us: “We were taught that reading is the ability to understand shapes on a piece of paper. But that’s like saying work is sitting in front of a computer. Technically it is, but it’s not the work. Reading is the ability to understand the material. Until recently, understanding the material depended on the ability to understand shapes on a paper. If you were dyslexic or hard of seeing, you couldn’t do it. I’m dyslexic, and a prefer audiobooks to print books any day of the week. I’m not stupid, but by the time I put in the time to read what’s written, I don’t have any energy left to understand what it says.”

EyeCam isn’t just complemental, but supplemental. It enhances its users’ senses and capabilities with AI, which in a sense, mind the pun, noninvasively turns them into demi-bionic people. “With language and text analysis and nimble understanding capabilities, we can have the hard of seeing get things even seeing people find hard to get,” says Wexler, illustrating with a feasible future feature: “When a user looks at a menu in a restaurant, he could tell the EyeCam ‘I want something dairy,’ ‘I feel like a hamburger,’ or ‘I’m vegetarian,’ and it’ll only read you the relevant dishes – faster than a regular person, who has to go through the entire menu.”

While AI is a mainstay of EyeCam, Wexler, only half-jokingly, casts doubt on its very definition: “This domain of AI is not well defined – we know what A is, but what is the I? There’s no definitive definition of intelligence. It’s not a distinct thing. Let’s take the Western Wall – it’s pretty intelligent, it lasted 2,000 years. Maybe it’s more intelligent than humans who live a 100 years. We’re not even close to fulfilling our human capabilities. AI is another step in that direction. The tractor helped, too. Those are tools that help humanity reach farther, and we need to understand how to leverage it. If a kid can’t understand shapes on paper, it’s a shame. You just lost this kid. He’ll tell himself he’s stupid. But if you assist him, you haven’t replaced his brain – it’s just a better way to hand him the information. And if he’s smart, he’ll take that information and be even more successful.”

Should we as humanity fully embrace AI, or hedge its potential dangers?

“Once things are easy to do, they need to be dealt with cleverly. AI is a fantastic tool. It can reduce road accidents, make disease discovery faster, cheaper and more accurate, and allow blind people to read books better than some sighted people. Such achievements undoubtably propel humanity forward and should be embraced. The faster we understand the added value of technology, the better it is.”

Wexler is optimistic when it comes to humans as well: “Machines can be built to keep working. Some of their work can be very fast. The combination allows machines to process more data that is possible by humans. The processing itself can be quite simple. We all know that computers can do basic math faster and with less mistakes than humans. General inference, and the ability to understand situations, are still very far off. We humans are amazing and we should never forget that. We are smarter than anything we create.”

[maxbutton id=”1″ ]