Bloomberg Businessweek reported on Aug. 5 that a 19-year-old Chinese teenager founded an AI (artificial intelligence) company in Silicon Valley in the U.S., and the company's market value has already exceeded $1 billion (roughly Rs. 7.023 billion).
The three-year-old startup Scale AI Inc. has been working to improve the state of the art in artificial intelligence, a technology that represents both humans and machines.Scale has built a suite of software tools that tag images before handing them over to a networked system of about 30,000 contract workers on the other end of the network to do the final touch-up work.
The scale of the technology has attracted many big-name customers in the self-driving car space, including Alphabet Inc.'s Waymo, General Motors Co.'s Cruise and Uber Technologies Inc.
Today, Scale is looking to sell its products to just about any company developing artificial intelligence technology. Several prominent venture capitalists have backed Scale, which plans to unveil an investment of more than $1 billion (Rs. 7,023 crore) on August 5th.
Alexandr Wang, co-founder and chief executive of Scale, said:
"It takes billions, if not tens of billions, of artificial cases of big data support for AI systems to reach the human level, and at present, only a few big companies can afford such a huge data collection program. "
Even by Silicon Valley standards, Wang is considered a genius. He grew up in New Mexico with parents who were both physicists. He participated in a number of internationally recognized programming competitions when he was a teenager and got job offers from tech companies while still in high school. This allowed him to graduate early, work in Silicon Valley, and then start his own business at age 19. Today, the 22-year-old Wang has $100 million (Rs 702.3 million) from investors, including Mike Volpi, a general partner at Index Ventures.
Mike said, "When we finished signing the letter of intent to invest, I ordered a nice bottle of wine to celebrate, and Wang was so young that I even had to ask him if I would be breaking the law by bringing him alcohol." (Wang was of legal drinking age by then.)
As companies race to build artificial intelligence systems on par with the likes of Google and Facebook Inc. they face two major challenges. One is getting enough data to support the machines in coming up with the final results. The other is ensuring quality data and results. While the machine can do a lot of work quickly, it does need people to be able to provide quality photos, text and video as a guide before it can come up with the results people want.
In the self-driving car industry, major companies spend millions of dollars a year hiring people to give them customer feedback on photos for data organization. Typically, such workers see a picture pop up on a computer screen and start by recognizing the outlines of cars and categorizing them in software. Then they analyze the buildings, parking spaces, pedestrians, traffic lights and more around the vehicle.
A picture can take anywhere from 10 minutes to several hours to process, and crews must go through each object, point by point, in a single photo. There are millions of images that need to be scanned in this way, and that data is then fed back into the AI system, and with that kind of intelligence, the feedback image recognition can be done quickly afterward.
Scale has built software that organizes such images and, in most cases, automatically labels all the objects that appear in them. Says Wang: "What used to take hours to accomplish now takes minutes."
Scale employs about 100 people at its San Francisco headquarters, in addition to an army of contractors around the world who work on image marking.Contractors are instructed in detail on Scale to complete specific data-gathering tasks, and the company continues to develop its software to create better intelligent super systems.
Wong did not specify where the contract workers come from or how much they make, but he insisted they are paid well. He said Scale would not try to optimize labor costs.