In 2018, blockchain and artificial intelligence continue to be two of the technologies generating the most buzz—and yet, for the former, not all of that buzz was positive. Blockchain, especially cryptocurrencies like bitcoin, were battered heavily this past year. Bitcoin had topped $17,000 early in 2018 and is now worth less than a quarter of that. That is the bad news.
The good news, however, is that the trend for AI is much more positive, with the technology gaining significant momentum consistently for the last few years. Established tech companies, governments, and cutting-edge startups are all looking for a slice of this growing pie. The government of Malta has created a task force focused on a national strategy for the technology and is building an AI citizen test for robots, for example. Apple has acquired four AI companies in the last two years, and IBM hosted its first-ever AI DevCon event.
Smaller, newer companies are gaining steam too, though. Leading Chinese startup Toutiao, which uses AI to recommend content to readers, inked a deal with Buzzfeed. And CherryHome, a home security AI system created by startup Cherry Labs, successfully raised $5 million and is on track to make its product available for purchase at the beginning of 2019.
Cherry Labs was founded by Nick Davidov Max Goncharov and Stas Veretennikov. Davidov also co-founded Cagarin Capital and served as an advisor to MSQRD. MSQRD was acquired by Facebook, and Fab.by, which was then in turn, acquired by Google. Goncharov is a co-founder of AR company 33Bits and is the winner of the TechCrunch Disrupt 2016 Hackathon. Both explained that AI’s potential is so great because it can be applied to such a wide range of use cases. Stas Veretennikov is ex-lead developer at Yandex, one of the most popular search engines in Easter Europe.
“AI is a general purpose technology—a type of computing with very versatile uses,” Davidov said. “Humanity has explored only a few of those so far and yet PriceWaterhouseCoopers names at least 300 current uses of machine learning, deep learning, and AI across different industries.” Additionally, according to JPMorgan research, “AI will contribute more than a quarter of all global economic growth in the next five years.”
Goncharov was also adamant about the fact that AI is only getting started, “As it goes from cloud to on-the-edge computing, it will get cheaper and cheaper,” he said, “To that end, machine learning will even impact low-tech devices. It will be ubiquitous.”
“That’s a huge difference from cryptocurrencies,” Goncharov said. “Cryptocurrencies didn’t require any strong technical or math revolution. But AI is growing based on new math discoveries and more and more powerful computing power.”
Cherry Labs employs 15 AI scientists with PhDs and, despite the scare-mongering that sometimes takes place about the future of AI, their aim is positive and simple: make people feel safer at home. So how can AI do that? In the case of CherryHome, the device uses computer vision algorithms to turn optical data—i.e. video footage that never leaves the device—into virtual skeletons in order to detect anomalies.
With the aforementioned $5.2 million funding round, which comes courtesy of GSR Ventures (a $1.9 billion fund), CherryHome will be partnering with elderly care agencies to provide smarter, round-the-clock care. This is a prime example of how AI inside the home can make real people safer and can help real professionals do their jobs better. The number of seniors in America is growing rapidly, and most want to live at home even as they age. But the reality is that aging people need supervision. They are at a greater risk for falls and they need medications and care—which is extremely expensive.
The AI on the CherryHome device can monitor whether an elderly goes into the bathroom and does not return, if they fall, or if their gait is abnormal. To protect the patient’s privacy, CherryHome turns them into a virtual skeleton and sends caregivers and family members real-time notifications of such anomalies. Also, all video footage is processed on-device—not sent to the cloud, as is the case with most home assistants. Already in place is a pilot partnership between CherryHome, TheraCare, in-home caregiving service and TriCura, a tech ecosystem for care agencies.
This represents another differentiator for AI, according to Goncharov. A lot of scientists in the AI space are working on fundamental problems—elderly care being just one of them. Looking forward, Goncharov says that AI will be further propelled as machine learning can be done with less and less data. The biggest hurdle to broader applications right now, he says, is the immense amount of data required to teach machines anything—another way that CherryHome is leading the way. CherryHome’s proprietary AI framework is based on a novel “nested autoencoders” approach, where devices learn from each other without human supervision and without needing giant datasets.
All in all, the successful application of AI in areas as crucial as healthcare and government are promising indicators that the buzzword is far from a fad. The buzz is real—and AI is just getting started.
Date: December 26, 2018