Why Safeway grocery clerks worry about artificial intelligence

March 17, 2020, 3:43 PM UTC

This is the web version of Eye on A.I., Fortune’s weekly newsletter covering artificial intelligence and business. To get it delivered weekly to your in-box, sign up here.

Business executives may be excited about artificial intelligence increasing productivity, but that doesn’t mean that rank-and-file workers are. 

Consider the grocery clerks at two Safeway stores in the San Francisco Bay Area. A few weeks ago, over 200 workers who are members of the United Food and Commercial Workers Local 5 (UFCW5) union picketed a Safeway store in San Jose, Calif. to voice concerns about a push by parent company Albertsons to add more A.I to its operations.

The picketing was related to the union’s recent contract negotiations with Safeway about improving pay and pension plans, among other issues, said UFCW5 director of strategic campaigns Jim Araby. During the final week of negotiations, grocery clerks complained to the union that Albertsons’ recent tests of A.I. at so-called micro-fulfillment centers at two Safeway stores in San Jose and South San Francisco would harm workers, Araby said.

Albertsons recently partnered with the startup Takeoff Technologies to create mini warehouses where computer vision technology automatically sorts items that shoppers order online. Using A.I. reduces the need for Safeway staff to manually locate and grab items for delivery—workers now just retrieve the finalized orders from a conveyor belt and sign off on them for eventual delivery.

Several grocery store chains are investing heavily in micro-fulfillment centers after Amazon helped to popularize as-fast-as-you-can deliveries, said Andrew Lipsman, a principal analyst at research firm eMarketer.

Araby explained that the union is less concerned about the micro-fulfillment centers obliterating jobs. Instead, the union worries that as Safeway adds more micro-fulfillment centers, higher-paid grocery clerks will be downgraded to lower-paid roles at the micro-fulfillment centers, which are part of the grocery store’s e-commerce unit.

“If more jobs go warehouse-style and they move the jobs there and pay them less, then it’s a huge problem,” Araby said. 

Safeway and Albertsons did not respond to requests for comment.

The UFCW5’s concerns echo those of other unions representing hotel staff and casino workers that view A.I. as inevitable and something that can’t be stopped. Rather than fighting against the technology, the unions are seeking protections in their contracts to require that employers train staff in newer skills or ensure that wages don’t plummet. 

Because the concerns about the micro-fulfillment centers came up so late in the negotiation process, the union didn’t address them in their tentative agreement with Safeway management last week, Araby said. The next time the union negotiates a new contract, likely next fall, it will bring up its concerns, he said. 

Although it’s too early to say what the union will seek, Araby imagines it could involve creating a “a joint labor-management committee” that would meet quarterly to discuss the company’s plans for implementing new technology and how it impacts jobs.

Araby acknowledged that the coronavirus pandemic could cause an “inflection point” that leads to more consumers using Safeway’s e-commerce and delivery service. But even if that doesn’t happen, he said, it’s clear grocery chains will continue to invest heavily in micro-fulfillment operations because “this is going to be part of the grocery income center.”

Jonathan Vanian 


The White House wants A.I. help. The White House Office of Science and Technology Policy along with representatives from Microsoft, the Allen Institute for AI, the Chan Zuckerberg Initiative, and others revealed a dataset containing 29,000 scientific articles related to the coronavirus that they intend for A.I. researchers to analyze on behalf of the public and the science community. “Now, The White House joins these institutions in issuing a call to action to the Nation’s artificial intelligence experts to develop new text and data mining techniques that can help the science community answer high-priority scientific questions related to COVID-19,” the White House said in a statement.

Gobble gobble. Tech giants like Apple, GoogleMicrosoft, Facebook, Intel, and Amazon are the biggest acquirers of A.I. startups, Bloomberg News reported based on data from research firm CB Insights. The article raises questions about the long-term ramifications of the bulk of A.I. startups being sold to only a handful of tech companies. “If big tech companies buy them all up, they eliminate these future competitors, and have a chance of actually owning the winners,” Primer AI CEO Sean Gourley told the publication.

An A.I. lawsuit involving misappropriated trade secrets. Facebook has been sued by A.I. startup Neural Magic for allegations relating to stealing the startup’s proprietary software and using it in its own A.I. projects after hiring one of its employees. The startup’s software helps researchers build neural networks—software that learns—that can work more efficiently with standard computer processors (CPUs) instead of the graphics processing units (GPUs) that many researchers typically use. Facebook CTO Mike Schroepfer previously told Fortune that the social networking giant uses large fleets of CPUs for some A.I. training tasks like “click predicting,” explaining that sometimes using multiple CPUs can be more efficient than even the most powerful GPUs.

That’s one way to lure retailers. Amazon is considering making some of the software used to power its Go cashierless stores available for free in an open source model, The Wall Street Journal reported, citing unnamed sources. The goal is to “gain favor” with retailers like Walmart that typically view Amazon as a competitive threat, the report said.  


ServiceNow hired Vijay Narayanan to be the enterprise software company’s first chief A.I. officer who will lead its advanced technology group. Narayanan was previously the head of content and discovery engineering at Pinterest and a Microsoft director of machine learning algorithms and data science.

Eric Horvitz is now Microsoft’s chief scientific officer while Peter Lee will be the tech giant’s research chief, CNBC reported. Horvitz has been Microsoft’s principal researcher for 16 years and is the co-chair of the company’s artificial intelligence and ethics in engineering and researcher committee.

Bill Gates has stepped down from Microsoft’s board, marking the first time he will have no formal role with the company he co-founded in 1975 with the late Paul Allen. Gates will be a technical advisor to CEO Satya Nadella.


Deep learning to determine who has the coronavirus. Researchers from institutions like Tel-Aviv University, Mount Sinai Hospital, and the University of Maryland School of Medicine published a paper about using deep learning to detect the coronavirus in patients and “can differentiate coronavirus patients from those who do not have the disease.” The researchers wrote that their system can “achieve high accuracy in detection of Coronavirus.”

Still, they caution: “In the case of a new disease, such as the coronavirus, datasets are just now being identified and annotated. There are very limited data sources as well as limited expertise in labeling the data specific to this new strain of the virus in humans. Accordingly, it is not clear that there are enough examples to achieve clinically meaningful learning at this early stage of data collection despite the increasingly critical importance of this software, especially given fears of a pandemic.”

Deep learning to track COVID-19 lung infections. Researchers from the Department of Radiology at the Shanghai Public Health Clinical Center and other institutions in China published a paper about using deep learning to analyze COVID-19 infections patients via chest CT scans. The paper said it’s A.I. system was “trained using 249 COVID-19 patients, and validated using 300 new COVID-19 patients.” The paper’s authors wrote that there are still several limitations to their study, including the fact that the A.I. system was trained on data that “may not be representative of all COVID-19 patients in other geographic areas.” But ultimately, they hope that the paper “could further reveal insights about imaging markers and findings towards improved diagnosis and treatment for COVID-19.”


IBM’s debating A.I. is added to Watson—By Jeremy Kahn

Despite fears of gridlock, Internet is holding up under coronavirus strain—By Aaron Pressman 

The best Twitter accounts to follow for reliable information on the coronavirus outbreak—By McKenna Moore

‘Where’s the leadership?’ A Q&A with WHO special envoy David Nabarro on COVID-19—By Erika Fry and Sy Mukherjee


Opportunistic A.I. As a journalist, I’ve become jaded to companies hungry for media coverage attempt to capitalize on major news events as a way to tout the benefits of their products. Regarding the rise of the coronavirus, I’ve already received several pitches from companies claiming their technology or services can lead to breakthrough treatments for the deadly virus. MIT Technology Review probes some of these claims and finds that “The hype outstrips the reality.” Although A.I. could help researchers develop better treatments in the future, it will “not save us from the coronavirus—certainly not this time.”

From the article:

“In fact, the narrative that has appeared in many news reports and breathless press releases—that AI is a powerful new weapon against diseases—is only partly true and risks becoming counterproductive. For example, too much confidence in AI’s capabilities could lead to ill-informed decisions that funnel public money to unproven AI companies at the expense of proven interventions such as drug programs.”

Read More

CEO DailyCFO DailyBroadsheetData SheetTerm Sheet