Chan Zuckerberg makes jump into artificial intelligence with acquisition of Meta
Monday, January 30, 2017
Facebook founder Mark Zuckerberg and his wife, pediatrician Priscilla Chan, have agreed to acquire Canadian startup Meta through their Chan Zuckerberg Initiative (CZI). The business and philanthropy team said they plan to make the capabilities of Meta’s artificial intelligence predictive modeling available free to researchers worldwide.
The acquisition is in support of CZI’s stated goal to make big bets in scientific initiatives to cure, prevent or manage all diseases by the end of the century.
Meta is a portfolio company of iGan Partners. According to Alexandre Nossovskoi, an analyst at iGan Partners in Toronto, which invests in technology companies, Meta’s growing expertise in big data applications accounts for the good fit with CZI. The Meta platform deploys artificial intelligence to read the world’s scientific papers. Its algorithm then interprets and directs that information to working researchers in real time.
This is no small feat. According to the CZI acquisition announcement, made by two senior leaders, researchers publish more than 4,000 scientific papers every day.
“Many of these papers will not be read by the scientists who could learn the most from them,” posted CZI’s President of Science Cori Bargmann and Chief Technology Officer Brian Pinkerton. “Scientists know that existing search tools can’t capture all of the relevant knowledge in this immense volume of scientific research. Meta is a tool that helps fill that gap.”
According to Nossovskoi, iGan began investing in Meta in 2013 because of its expertise in big data. He said that before Meta, the standard search options for scientists to try to keep up with available research, such as Google Scholar, were very inefficient.
“I’m sure any researcher will tell you how much they hate it,” Nossovskoi told CenterWatch. “Any startup we look for has that commodity—big data. For us, that’s what we see as the future.”
Such big data applications—also known as deep learning—is also making its way into big pharma.
“We are applying deep learning in biomarker development and drug discovery and drug repurposing,” said Dr. Qingsong Zhu, COO at InSilico Medicine. “Our algorithms can effectively predict drug/compounds efficacy and forecast clinical trials outcomes. By applying generative adversarial networks, InSilico Medicine could generate new anti-cancer drug molecules on demand.”
He said that CZI’s plan to offer the Meta platform without charge amounted to “kindness and generosity” that will be of real benefit to researchers.
“After Meta joins CZI, its products will be free to the entire scientific community. It promotes the research and development and helps cure diseases,” said Dr. Zhu.
Dr. Zhu also said that big pharma is relatively slow to develop new technologies such as artificial intelligence. He hopes to see such progress in the pharmacy sector within the next two years.
“Although there are several big pharmas contacting InSilico Medicine and seeking collaboration in applying our AI technology in drug discovery, we still [do] not see a big movement, [such] as the acquisition of Meta, in the pharma industry,” said Zhu.
Such is not the case in most other medical and scientific sectors. The use of artificial intelligence is one of the hottest areas in medicine. Until recently, using big data in predictive modeling has been unmanageable due to the limitations of several technologies, including algorithms. But science and medical companies have crossed that barrier with new capabilities—and they are beginning to announce these developments in the marketplace.
For example, at the recent 2016 Radiological Society of North America (RSNA) annual meeting in Chicago, which had a registered attendance of 54,037, several companies rolled out artificial intelligence product developments. Some will be available as soon as this year in the medical field.
Among these is IBM’s Watson platform. The new Watson cognitive assistant offers a peer review tool that will sort through electronic medical records and flag cases of undiagnosed aortic stenosis, a life-threatening heart valve condition that affects 1.5 million people every year in the U.S. Watson’s artificial intelligence looks for clinical evidence when physician notes and data in the electronic health record do not match.
Alphabet’s DeepMind, a London startup which previously sat under Google but has now been spun out, is also in development of healthcare artificial intelligence. Although there have been a few wrinkles—such as the recent criticism for using patient information to develop a predictive app to help doctors treat patients with kidney and eye disease—the company has a well-known commitment to big data medical solutions.
For Nossovskoi, artificial intelligence will dictate the areas of emerging science.
“Publishing houses have to look at 20,000 papers a year to decide what to publish,” he said. “Meta can review millions of papers, constantly scanning the horizon for emerging sciences. For example, it could indicate to big pharma companies where to put their money in research and development.”
He added that for Canadian companies, the CZI acquisition of Meta is just the high-level exposure that the Canadian research and development sector needed.
“Canadians are cautious investors” said Nossovskoi. “Startups often have difficulty getting later stage funding.”
This article was reprinted from Volume 21, Issue 04, of CWWeekly, a leading clinical research industry newsletter providing expanded analysis on breaking news, study leads, trial results and more. Subscribe »