Fast food restaurants are exploring the possibility of using AI chatbots in drive-thrus. Given the improvements with speech recognition tech and machine-learning-based processing, it’s definitely a logical next step. It doesn’t require expensive mechanical automation – just the fast-food equivalent of Siri or Alexa, custom-trained on a language model appropriate for restaurants.
Based on what I’m reading, these AI chatbots aren’t quite “ready for prime-time.” What they’re doing, however, is creating a “competency floor.”
For example, the chatbots’ current order-taking accuracy is about 85%. Based on my years as a teenager working in fast food, I know that’s less than a highly-competent fast food worker – but it’s more than a minimally-competent worker that doesn’t care about their job and/or isn’t paying attention to what’s going on.
Order accuracy is important, and on a large scale a failure rate of 15% is unacceptable. But if a manager can flip a switch and activate an AI that has an 85% accuracy rate for order-taking, there’s no reason to ever employ somebody whose rate is less than 85%. That’s now the “competency floor.” It’s like the bar at amusement parks that says “you must be at least this tall to ride” – except instead of height, it’s competency.
And the bar is rising. While employees can be trained and motivated, that typically requires significant personal investment on a per-employee basis. Improving the AI is more expensive, but gives the benefit across the entire chain (or industry).
There’s still a long way to go, but I think this is going to be the trend. Jobs that largely involve transcribing customer input and asking easily-defined follow-up questions (“would you like fries with that?”) are on the way out. The question for humans facing that world is easy, but not simple to answer. What things can we do as humans qua humans that are hard to replicate with an AI?