.png){kind=logo}
Taco Bell is reportedly re-evaluating the next stage of its AI strategy after errors made by the company’s drive-through automated voice system went viral.
The restaurant chain’s chief digital and technology officer, Dane Matthews, admitted to the Wall Street Journal (WSJ) that the implementation of voice-based AI has not been without its challenges.
He added the company has learned “many lessons” from the AI rollout, including not to use it in drive-throughs.
Last year, the retailer rolled out AI voice technology across over 500 stores in the US.
The technology is used to simplify the ordering process and boost overall productivity during peak times.
But Taco Bell’s Matthews admitted to the WSJ: “There are times when humans are better suited to taking orders, especially when restaurants are busy.”
The statement comes as number of videos on social media made fun of glitches and mistakes produced by the AI voice system as customers tried to order food.One video shows a customer who appears to crash the system by ordering 18,000 glasses of water.
In another video, a person becomes increasingly annoyed as the AI keeps asking if they would like to add drinks to their order.
In a video posted on Instagram, which has been viewed over 21.5 million times, a man orders “a large Mountain Dew”, but the AI voice keeps asking “and what would you like with that?”, prompting the man to punch the steering wheel in frustration.
Taco Bell said that the videos posted online do not do the technology’s success justice.
According to the company, over two million orders have been successfully processed since introducing the AI voice system.
However, the chief tech officer has stated that the company is now taking a different direction and will not leave AI unchecked.
“We’ll help coach teams on when to use voice AI and when it’s better to monitor or step in,” he added.
