Big Tech’s research teams on shaky ground

The research arms of Big Tech companies have made breakthroughs galore — but as the industry’s role in society grows dominant and researchers examine tech’s human dimensions, corporate labs are also becoming lightning rods of controversy.

The big picture: Academic researchers claim the freedom to pursue their studies wherever they lead. Corporate research departments profess independence, too — but that ideal can face tension with a company’s goals and profit-seeking.

Driving the news: These issues have resurfaced as part of a series of controversies within Google’s AI research team, including the recent forced exits of Timnit Gebru and Margaret Mitchell, the co-leaders of Google’s Ethical AI team.

• A part of the dispute has focused on the freedom of researchers within a company to publish their research unfettered and to speak freely about the shortcomings of their employer’s own products.
• “Research integrity can no longer be taken for granted in Google’s corporate research environment, and Dr. Gebru’s firing has overthrown a working understanding of what kind of research Google will permit,” a letter signed by thousands of Google employeesand others academics and techies outside the company said.
• To be clear, the Google controversy is about more than just academic freedom — critics are also pointing to an array of concerns about how the company treats women and people of color.

Be smart: Exploring semiconductor materials or battery design is less likely to raise red flags in corporate suites than studies of algorithm fairness or technology’s impact on workers’ health.

• At the same time, company-funded research on such topics is less likely to earn public trust, no matter how impeccable the scholar’s work.

Each of today’s tech giants takes a differing approaches to research.

• Microsoft Research dates back 30 years and now has thousands of researchers across labs in China, India, the U.K. and across North America. Among the products that started in research are the Xcloud gaming service and the Airbandbroadband initiative, as well as significant features in Excel and other Office apps.
• Google employs more than 3,000 researchers, with labs in the Bay Area, New York City, India, Zurichand London. Among the commercial products that started life within research are Google Translate and Tensor Flow, as well as many of the photography advances in the Pixel smartphone and the smart reply feature in Gmail.
• Apple, while known for keeping its research secret and largely focused on future products, has taken a slightly more open approach when it comes to machine learning, given that the field remains dominated by academic types. A series of academic papers, for example, are posted on the company’s public web site.
• Amazon is most known for taking a product-centered (or, as Amazon would call it, customer-obsessed) view of R&D. However, it does have an Amazon Scienceunit doing research work in a number of fields such as computer vision, machine learning and quantum computing.

History lesson: 20th-century corporate research labs ushered in the modern technology era.

• Transistors were born at the now 90-year-old Bell Labs, which today is part of Nokia. Bell also birthed the laser, Unix and solar cells.
• Founded in 1970, Xerox’s Palo Alto Research Center(now Parc) helped invent and develop fundamental computing building blocks including the graphical interface, Ethernet networking and laser printing.

Our thought bubble: In the past, research divisions often became the first target for budget cuts at giant corporations that loved to put Nobel Prizes in press releases but struggled to turn breakthrough discoveries into profitable products.

• Today’s tech giants aren’t cutting budgets, but may still be tempted to pull back from research if they can’t control it.

Πηγή: axios.com