May 18, 2024

The fast food landscape is evolving with the introduction of artificial intelligence at drive-thru windows. Utilizing technology developed by Presto and incorporating elements of OpenAI’s GPT, AI chatbots have begun taking orders across various fast food chains, including Hardee’s, Del Taco, and Checkers. This technology uses automatic speech recognition to transform spoken orders into text that the AI system can process. The Wall Street Journal’s Joanna Stern recently tested these AI chatbots at a Hardee’s location to assess their functionality and efficiency in real-world conditions.

Testing the Technology
Stern’s examination of the AI-driven order system involved several challenging scenarios to evaluate the technology’s robustness and adaptability. Through repeated visits, she presented the AI with various speech recognition challenges, including placing orders amidst loud noises like car horns and barking dogs. The AI chatbot’s response to these interruptions was critical in understanding its capability to function in typical noisy environments occurring daily at many drive-thru locations.

Speech Recognition Capabilities
The first phase of Stern’s testing focused on the chatbot’s ability to convert spoken words into text accurately. This is essential for the AI to interpret the customer’s order correctly. Stern ordered items like the Big Hardee Combo, modifying the order with requests for extra ketchup and a Diet Coke to see how well the AI could handle specific and detailed instructions. The chatbot’s performance was generally good, recognizing and confirming the order details despite disruptive noises, a testament to its sophisticated speech recognition technology.

Challenges with Background Noise
Despite the AI’s adept handling of straightforward speech recognition, Stern’s tests revealed that the system is not impervious to errors introduced by ambient noise, a common feature of drive-thru environments. For example, when ordering chicken nuggets with the sound of barking dogs in the background, the chatbot asked about the choice of dipping sauce but struggled with follow-up questions, showcasing limitations in handling unexpected auditory disturbances.

Understanding and Processing Orders
After converting speech to text, the next step involves the AI’s understanding of the order through its Natural Language Understanding module. This system needs to grasp the basic request and anticipate and manage follow-up queries related to the order. Stern’s interaction included changing an order partway through. She initially asked for a Big Hardee Combo but then switched to the Superstar with cheese. The AI successfully navigated this change, suggesting it can handle typical customer modifications without confusion.

Complex Queries and Human Intervention
However, the AI system showed shortcomings when dealing with complex customer questions beyond simple order modifications. Inquiries about ingredients or calorie content often stumped the AI, leading to a prompt for human assistance. This indicates that while AI can manage routine orders effectively, it still relies on human employees to handle more detailed or nuanced customer interactions.

The Role of AI in Customer Service
Integrating AI into drive-thru service points towards a future where fast food operations can become more efficient while relying on human oversight for quality control and complex decision-making. Stern’s tests suggest that AI can enhance service speed and accuracy but also highlight the need for ongoing human involvement, especially for quality assurance and customer satisfaction.

As AI technology continues to evolve, its potential to support the fast food industry grows. However, the tests conducted by Stern illustrate that while AI can significantly streamline some aspects of the service process, complete autonomy without human backup is still far off. The balance between automated efficiency and human judgment remains crucial, ensuring that as AI takes on more roles in the industry, it enhances rather than detracts from the overall customer experience.