New York CNN  —  There may soon be one fewer person standing between you and your Cheesy Gordita Crunch. Taco Bell is set to expand the use of artificial intelligence voice technology in its drive-thrus to hundreds of US locations by the end of this year, parent company Yum! Brands said Wednesday. The move means that when users pull up to place their order at a Taco Bell, there’s a good chance they will be talking to a computer, rather than a human employee, on the other end. Taco Bell isn’t the only fast-food chain to test having customers talk to AI in its drive-thrus in an effort to free up employees to spend more time on other tasks and boost sales. But Taco Bell’s expansion comes after a prominent rival had trouble trying to use similar technology. A 'Drive-Thru' lane at a McDonald's Corp. fast food restaurant in Louisville, Kentucky, U.S., on Wednesday, Oct. 20, 2021. Luke Sharrett/Bloomberg/Getty Images Related article McDonald’s pulls AI ordering from drive-thrus — for now McDonald’s said last month it was pulling the plug on the AI ordering technology that it was testing at more than 100 US drive-thrus, after customers complained about the system getting their orders wrong. Viral videos showed customers ending up with absurd orders as a result of the McDonald’s AI drive-thru technology, like the woman who didn’t really want nine sweet teas or the girl shouting “stop” at the screen as it tallies up more than two dozen orders of chicken nuggets. That AI struggled to perform well in a drive-thru setting isn’t totally surprising. The technology needs to be trained on huge sets of voice samples to correctly understand the full range of human accents and speech patterns, and noisy environments can throw off AI speech recognition tools. And we all know the frustrating experience of being on the other end of a poorly functioning computer assistant, desperately wishing we could just talk to a human. But Lawrence Kim, chief innovation officer at Yum! Brands, told CNN he’s confident that his company’s technology — which is already in use at 100 Taco Bell locations across 13 states — won’t have the same issues. Instead, he said, the technology has actually led to greater ordering accuracy, as well as happier employees and shorter drive-thru wait times. “We are confident that we’ve approached this the right way,” Kim said in an interview, adding that it’s taken two years of testing, gathering employee and customer feedback and improving the technology to get to this point. Yum! Brands did not provide data on the improvements it said it’s seen thanks to the AI ordering, beyond noting that it now earns $30 billion in sales (about 50% of its total revenue) from digital-first ordering channels. Kim declined to specify which technology partners Yum! has worked with to develop the AI ordering tool. Kay-sah-dee-ya, Kay-sah-dill-uh One thing the company has learned along the way: Not every customer uses the same words to order a specific item, nor pronounces them the same way. For example, Kim said, the company had to train its AI model to understand not only the correct pronunciation of “quesadilla” (kay-sah-dee-ya) but also the incorrect, but still common, “kay-sah-DILL-uh.” And if there are issues or confusion, a Taco Bell employee is still always listening on the other end of the ordering system with the ability to intervene. Kim also denied that AI ordering technology would replace human jobs, a growing concern around all manner of new AI tools. “Voice AI does not replace any team members,” Kim said, adding that the technology is instead designed to “enhance the team member experience so they can focus on other tasks that are a priority for them.” Still, Kim said he does think AI could ultimately be implemented into other areas at Taco Bell and the company’s other chains, which also include Pizza Hut, KFC and The Habit Burger Grill. “I believe the applications for voice AI are boundless,” Kim said. “The technology is emerging and evolving so very quickly that we … as an innovation team as well as a technology team are continuously evaluating what could be next.”