Artificial intelligence continues to revolutionize various fields, but recent hiccups serve as a reminder of its ongoing development. McDonald's decision to halt its AI drive-thru voice ordering system after social media exposure of its errors exemplifies this. While the technology holds immense promise, real-world challenges remain.
McDonald's USA Chief Restaurant Officer Mason Smoot told franchisees in a memo obtained by CNBC, "After a thoughtful review, McDonald's has decided to end our current partnership with IBM on AOT (automated order taking) and the technology will be shut off in all restaurants currently testing it no later than July 26, 2024."
In 2021, the fast food giant partnered with IBM to test-run the AI ordering technology at over 100 McDonald's locations.
"While there have been successes to date, we feel there is an opportunity to explore voice ordering solutions more broadly,” Smoot's memo said, per CNBC. “After thoughtful review, McDonald's has decided to end our current partnership with IBM on AOT. IBM remains a trusted partner and we will continue to utilize many of their other products across our system.”
McDonald's began testing AI drive-thru ordering to "determine if an automated voice ordering solution could simplify operations for crew and create a faster, improved experience," according to a statement obtained by Fox Business.
While McDonald's hasn't officially disclosed the specific reasons behind ending its AI drive-thru test, reports suggest accuracy may have played a significant role. According to Fortune Magazine, social media has been abuzz with customer complaints about receiving incorrect orders when interacting with the AI system. These reported inconsistencies likely contributed to the decision to shelve the project for now.
Following McDonald's announcement of ending its AI drive-thru test developed with IBM, discussions have arisen regarding the contributing factors. Some speculate that technical limitations of the AI system itself might have played a role. Others suggest that integration challenges or limitations in how McDonald's implemented the technology could have been a factor.
A user wrote on X, "I think AI is great and it should be used, but it is terrible when interacting with humans when humans want something very specific. Like a chatbot that tries to answer a question a human has on a website or AI trying to take an order for specific things, we aren't there.”
Another user wrote, "Mcdonald's should have hired four small vendors to try to solve the problem, kept whichever one worked best, and probably still spent less."
The third user commented, "More likely that both McDonalds and IBM share the blame. MCD's digital execution has been a mess all around. Have you tried using their app?"
“In the past, IBM has severely mishandled Watson for hospitals. The bigger question is, how did they secure a deal? Their AI technology is poor. It must be due to some sort of sales relationship with top management rather than an objective tech evaluation,” the fourth user wrote.
