This is the second post of TechNode’s profile of AI and robotics firms present at the Beijing edition of Global Mobile Internet Conference or GMIC, which opened with a warning from Stephen Hawking that AI could end humanity. Read Part 1 here.
A good number of “humanity-ending” AI firms and robots were present at GMIC, where the robots zoomed around or performed energetic aerobics on the exhibition floor. TechNode continues our profile of these companies and their thoughts about the potential AI apocalypse.
BooCax is a supplier to robotics firm, providing self-positioning robot navigation technology that is second to none in China. BooCax combines existing lidar with proprietary ultra-wideband technology to equip robots with superior self-positioning capability.
“The robot can be anywhere inside and it won’t lose its own position on the map,” BooCax sales manager Xiao Xuesong explained. “So far, this is an issue that other companies in the industry have not solved.” The accuracy of the self-positioning system is +/ – 20cm.
Founded by a team of researchers from Stanford, Tsinghua and the Chinese Academy of Science, the firm is also offering its own robot, BamBoo. Targeting banks, hotels and other hospitality businesses, the robot is a service-type robot that can handle objects and basic interactions.
What happens when AI becomes self-evolving?
“Then it’ll be like the movies,” Xiao said.
“Um. Hopefully not.”
Founded by former Microsoft Partner Engineering Director Kenny Chien in 2015, Emotibot aims to create the first artificial intelligence technology that recognizes emotion from facial and voice recognition as well as textual analysis (read our interview with Kenny Chien here).
Currently, Emotibot is used by clients such as Ctrip, JD Finance, iQiyi for their customer service chatbots. Chien describes Emotibot as Alexa but Chinese, as the bot recognizes intent, meaning and emotion. The goal for Emotibot is to become self-evolving and capable of feeling emotions like the artificial intelligence personality Samantha featured in Spike Jonze’s movie Her.
At GMIC, Emotibot demonstrated its facial recognition technology. It correctly identified the writer as female, adult, wearing glasses, has black hair and seems to be happy (due to the smile). However, it also had the questionable metric of attractiveness, which the writer only scored a 72%. Maybe artificial intelligence could really be the worst thing to happen to humanity.
Is AI the worst thing to happen to humanity?
“In my opinion, for the next few decades, artificial intelligence will be a good thing for humanity,” Emotibot product manager Yang Liang predicted. “But Hawking must be looking really far out, beyond the next few decades, when AI development has reached an unimaginable crescendo. Then it’ll be really hard to tell.”
The Silicon Valley startup Slightech is known more for its Mynt tracker, which is billed as world’s thinnest and most powerful smart tracker and has won a Red Dot design award. However, the product showcased at GMIC was their newest Mynt robot, known as Xiaolan in Chinese. The small arms featured in its design are reminiscent of EVE, the sleek white robot from Wall-E.
Slightech was founded by Dr. Leo Pang, a Stanford computer science graduate who has worked in research prior to starting the company. In 2015, Slightech released the Mynt robot which it claims to be “arguably the world’s most powerful robot”. It has real-time facial and voice recognition and facial triggered motion, as well as machine learning capabilities.
Currently, Mynt robot is targeted mostly at B2B clients. Since being launched, around 40 Mynt robots have been sold. Sales manager Kyle Pang thinks that 2020 will be the year that service-type robots will really gain an inroad into households.
What do you think of Hawking’s message?
“Currently, robots have taken care of a lot of dangerous tasks for humans, which is an achievement,” Slightech sales manager Kyle Pang (Leo Pang’s brother) said. “We have talked about this issue before, but robots cannot think for themselves at this stage. Humans can deceive robots. But for robots to deceive us, it’ll take a very long development. During this development, we should try to control and prevent the [robot uprising] as much as possible.”