|
|
Software Development May Provide Model for AI Regulation; Venture Funds Push Into Enterprise Tech; Google's Medical-Records Quest
|
|
|
|
|
|
Good day. Everyone knows that the current regulatory regime doesn't really work for artificial intelligence companies, one federal official tells WSJ Pro. The question is whether software development, with its focus on quick, incremental improvements over a long period of time, can work better, Carlo Martuscelli reports.
Also today: Venture funds have tired of pouring money into the consumer sector, and are financing less-capital-intensive startups focused on business technology. Read on.
|
|
|
|
|
IDx's AI-based device was the first approved by the FDA that diagnoses a medical condition without a doctor’s guidance. It uses algorithms to analyze images of retinas to detect signs of vision loss in patients with diabetes. PHOTO: IDX TECHNOLOGIES INC.
|
|
|
FDA official envisions new model for AI regulation, based on software development. In 2018, IDx Technologies Inc. became the first company to receive federal approval for an AI-based device that diagnoses a medical condition without a doctor's guidance. The company, based in Coralville, Iowa, worked with the U.S. Food and Drug Administration for eight years to get approval for the device that detects signs of vision loss in patients with diabetes.
One interesting twist from a regulatory perspective is that the company continues to work closely with its regulator. "We are always improving, and notify the FDA consistently in line with current guidance," IDx founder and Executive Chairman Michael Abramoff told Carlo Martuscelli for WSJ Pro. More devices have been approved since IDx’s, with others still under consideration.
The aim is to approach regulation more like how AI software itself is developed, responding to incremental improvements with continuous oversight, Mr. Martuscelli writes.
Bakul Patel, the director of the FDA's digital-health division, said FDA draft regulations released in April are an attempt to move away from the current system, where devices go through a one-time regulatory review and are revisited only in cases of severe failure.
“We all understand that the regulations don’t quite fit,” he said. “We need to evolve, and we need to work with stakeholders to make sure this paradigm...is done well.”
|
|
|
Google’s quest for medical records. Roughly a year ago, Google dispatched former CEO Eric Schmidt to pitch health-data company Cerner Corp. an unusually rich proposal, the Journal reports. In a series of calls, he offered around $250 million in discounts and incentives, people familiar with the matter say. Cerner was interviewing Silicon Valley giants to pick a storage provider for 250 million health records, one of the largest collections of U.S. patient data. Google representatives were vague about how Cerner’s data would be used, making Cerner executives wary, the people say. Eventually, Cerner struck a storage deal with Amazon instead.
|
|
The failed Cerner deal reveals an emerging challenge to Google’s move into health care: gaining the trust of health care partners and the public. So far, that has hardly slowed the search giant. Google has struck partnerships with some of the country’s largest hospital systems and most-renowned health-care providers, many of them vast in scope and few of their details previously reported. In just a few years, the company has achieved the ability to view or analyze tens of millions of patient health records in at least three-quarters of U.S. states, according to a Wall Street Journal analysis of contractual agreements.
|
|
|
|
Geneticists call on Myriad to share data. A leading medical society is calling on Myriad Genetics and other lab companies to share proprietary genetic testing data in a public database, the Journal reports. The call by the board of the American College of Medical Genetics and Genomics, a professional organization that also publishes influential guidelines on classifying genetic variants, came in response to a Dec. 20 Journal article that described how seven women in one family had surgery after learning from a Myriad test that they had a variant on a BRCA gene that significantly elevated their risk of hereditary breast and ovarian cancer.
|
|
|
|
$6.6 Billion
|
Estimated size of the health-care AI market by 2021, up from about $600 million in 2014, according to Accenture PLC.
|
|
|
|
|
|
TikTok signage displayed at an event hosted by Bytedance in Tokyo last year. Sequoia Capital made a $384 million investment in the owner of video-sharing app TikTok. PHOTO: SHIHO FUKADA/BLOOMBERG NEWS
|
|
|
Sequoia taking its time as it invests massive fund with enterprise focus. Venture-capital firm Sequoia Capital is in no rush to invest its $8 billion flagship fund, Yuliya Chernova and William Louch report for WSJ Pro. When Sequoia started looking for investments, it found that the era of substantial private financing of transportation, food-delivery, home-sharing and other consumer companies was largely over, and turned to less-capital-intensive enterprise software companies. The Menlo Park, Calif.-based venture fund has invested less than 20% of its Global Growth Fund III, the largest venture fund collected by a U.S. firm, in about 18 months, putting it slightly behind on its
intended five-year investment pace, according to people familiar with the firm’s strategy. It has shifted to smaller transactions, too.
|
|
Venture funding flows into the enterprise. The normal order of funding in venture capital flipped last year, Sarah McBride writes for Bloomberg. PitchBook data compiled for Bloomberg shows that enterprise tech companies that specialize in software or services for businesses raised $30.42 billion, about one-third more cash than consumer technology companies, Bloomberg says. Enterprise companies’ venture haul for 2019 was almost double the previous year’s, according to Bloomberg. Cash invested in consumer companies fell by almost a quarter between 2018 and 2019, to $23.26 billion, Bloomberg says, citing PitchBook data.
|
|
|
|
U.S. export controls may limit tech sector hubs in China. The Commerce Department imposed new export controls on AI, "a measure apparently designed to prevent US companies from shipping AI technology that could train Chinese military drones or teach intelligence systems to interpret aerial imagery," Wired reports. That could interfere with U.S. tech companies that have established research labs in China to tap talent and keep track of trends, according to Wired. “I don’t think any US government official expects to keep China from developing AI—they are doing quite well—but they don’t want companies like Google or Microsoft helping them,” says James Lewis, senior vice president of
the Center for Strategic & International Studies, a Washington, D.C., think tank.
|
|
|
|