...
By Anna Rahmanan
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
For years, the future of the restaurant arrived one touch screen at a time. You entered, tapped, customized, paid and waited for your number, while the counter—once the place where the restaurant met the customer—slowly began to resemble an unattended gate at an airport.
Now comes the strange part. Just as artificial intelligence is becoming capable of doing considerably more of the work involved in feeding us, some of America’s biggest restaurant chains are rediscovering an older technology—a human being saying hello.
McDonald’s, Burger King, Wendy’s and Starbucks have all put new emphasis on hospitality and customer experience after years of pushing ordering toward kiosks, apps and automated systems. Burger King said this week that some customers in its drive-through AI trials were leaving instead of talking to a bot. The old becomes new again, it seems.
Read More on Life
The temptation is to call this a backlash against technology…but what if what is really happening is actually the reverse? What if AI is actually making human service more valuable?
The Counter Becomes Premium Again
McDonald’s offers a pretty good example. This week, the company unveiled its new McDonald’s > NEXT strategy, which plans to deploy its generative-AI-enabled ArchIQ system at scale as part of an effort to improve restaurant efficiency. In the same plan, McDonald’s created a “People > NEXT” pillar designed to equip workers to deliver “a new level of hospitality,” along with a new companywide program called Make It Golden centered on food and hospitality.
In other words, the machine and the smile are arriving together.
It seems like the restaurant industry’s automation experiment has exposed a distinction that was easy to miss when technology was less capable. There is restaurant work, and there is hospitality. The first contains an enormous number of predictable tasks that computers can increasingly organize, accelerate or perform, perhaps even better than human beings. The second, though, begins exactly where predictability ends.
Related Story
Starbucks Won the Battle of Seattle. Did the Protesters Win the War?
Charan Nukathoti encountered that boundary while building Savor Scout, an AI-powered restaurant-discovery service. The product can digest menus and reviews to find restaurants matching what somebody wants. When allergies enter the conversation, however, Nukathoti deliberately sends the user back into the analog world.
“AI is very good at reading everything ever written about a restaurant,” Nukathoti told Newsweek. “It’s useless at knowing what’s true there today.”
A model can find a menu, but it cannot see what happened in the kitchen at lunch.
“No model knows that the fryer got shared last Tuesday or that the kitchen changed hands in March,” Nukathoti said. “The chains leaning back into human service aren't being nostalgic—they've correctly identified the part that doesn't automate.”
There is a broader lesson hiding inside the allergy example. AI’s great advantage is informational abundance. It can read more than any person could. Restaurants, though, are physical places filled with tiny, current facts—the oven running late, the soup is sold out, ingredient substitutions, table 12 waiting for too long. Hospitality is built from noticing those kinds of things.
The Best AI May Be the AI You Never Meet
Starbucks appears to have reached much the same conclusion from the opposite direction.
The company is simultaneously investing in AI and trying to make its coffeehouses feel more human. Its Green Apron Service program puts more emphasis on personal customer interactions, while its technology strategy increasingly places AI behind the bar rather than between the barista and the customer. Starbucks says tools such as Green Dot Assist can answer employee questions, while Smart Queue organizes orders across café, mobile, drive-through and delivery channels. The company describes its preferred role for AI in unusually revealing terms: “At its best, AI is invisible.” Hide the machinery!
Starbucks has some early reason to believe the combination can work. In July, the company reported a fourth consecutive quarter of positive global comparable-sales growth, while CEO Brian Niccol credited a strategy built partly around “human connection” and said Green Apron Service was helping employees deliver a more consistent experience.
...
Dan Jin, an assistant professor of hospitality and tourism management at Rochester Institute of Technology’s Saunders College of Business, has studied the division of labor between humans and machines. Her research has found that successful automation depends in part on clearly defining which roles belong to technology and which remain human.
“I would use a simple test: Is this a predictable task, or does someone need to understand an individual guest?” Jin told Newsweek.
Forecasting meals, routing orders and reordering a familiar lunch lend themselves to automation, she said. A serious allergy, a wrong meal or a long delay creates a different problem.
“A guest usually wants more than a fast response,” Jin said. “They want to know that someone understands the problem and can fix it.”
The important word there may be someone.
Nobody Actually Wants to Abolish the Kiosk
There is a danger of romanticizing all this. Plenty of customers do not want an artisanal conversation with a cashier before receiving their fries and, sometimes the highest form of hospitality is getting lunch into somebody’s hands in 90 seconds.
Automation can also make restaurants easier to use. A kiosk lets a customer browse without holding up a line, for example, and an app can remember an order. AI can also forecast demand and reduce the amount of chaos employees have to manage. McDonald’s itself is pairing its new hospitality push with a plan that it says can generate roughly 250 basis points of gross restaurant-level efficiency gains.
Burger King, meanwhile, is hardly abandoning AI. It has been testing voice AI in 1,500 drive-throughs, among other potential updates. What it is changing is the assumption that automation should become compulsory: the chain plans to make it easier for customers to reach a human after its pilots found that some people drove away rather than order from a bot.
Jin’s formulation is useful here because it avoids the increasingly tedious choice between techno-utopia and techno-rejection. The same customer who happily orders a weekday coffee from an app may want a server to notice an anniversary dinner.
“Automation becomes intrusive when it interrupts the guest’s goal,” she said. “The guest ends up working for the system.”
That may be the real restaurant revolt here: customers do not necessarily object to machines doing work, they object when the machine gives them more work to do.
Hospitality After the Machine
The likely future, then, is stranger than either the restaurant industry’s old automation dream or the current nostalgia for human service suggests.
Restaurants may become far more automated so they can feel less automated.
AI can sit underneath the experience: forecasting inventory, sequencing tickets, handling routine orders, spotting bottlenecks and searching through a lot of information. A person can occupy the increasingly valuable territory above it: judgment, reassurance, exception, recovery, recognition.
Nukathoti puts the dividing line between information that can be synthesized and reality that has to be verified. “In some cases,” he said, “the best AI experience may actually be one that gets you to the right human faster.”
That is a pretty old-fashioned destination for the newest technology.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines