AI continues to expand as its herd mentality drives nearly every corner of the tech world toward the shiny new thing. Some resulting AI use cases will sustain, while many won’t stick to the wall. And like any emerging tech, other killer apps will emerge that no one has thought of yet.
The latest subsector sprinting toward AI is the local on-demand economy. We’re seeing last-mile food delivery and micro-mobility apps start to integrate conversational AI to enhance their UX. This involves everything from automated checkouts to getting good sushi recommendations.
Over the past week, we’ve seen a few rumors, integrations, and product launches to that effect, which makes it a trend. Here are a few examples that characterize what we’re seeing.
Ep. 27 Maps Intersection of AI & Websites with Newfold CEO Sharon Rowlands
Code Breaking
Uber Eats has developed a conversational AI chatbot meant to provide food recommendations for users and streamline delivery orders. This was uncovered recently by developer Steve Moser (and reported by Bloomberg) who found hidden code within the Uber Eats app.
From what was gathered by Moser, the chatbot’s function seems to ask users probing questions that can train the agent to be an effective food assistant. Beyond training, the bot can be used in individual situations to quench a given urge, given prompts like budget and food types.
In addition to discovering food, the AI agent is also meant to streamline the ordering process, as noted. Think of this as an alternative UX that involves “conversations” rather than eCommerce checkout flows common to food-ordering apps. This leans into current demand signals.
Though Uber hasn’t confirmed the nature of the uncovered code, CEO Dara Khosrowshahi recently told Bloomberg Television that an AI chatbot was in development. So the found code seems to align with that. To be fair, Uber Eats already uses AI to match customers and drivers.
Accurate or Aggravating
DoorDash is meanwhile focused on a singular use case: handling inbound calls, Its new voice chatbot will handle orders and offer callers recommendations to complement their meals. This flows from Doordash data that a fifth of take-out diners prefer calls, while half go unanswered.
The point here is to serve restaurants more than users. DoorDash expressly wants to help them boost sales through AI-driven upsells. It will do this by training the AI on the most complimentary pairings, and therefore recommendations most likely to evoke mouthwatering moments.
The question is if consumers will bite, or if they’ll quickly get fed up. It’s no secret that there’s a universal aversion to automated voice response (AVR) systems. Perhaps OpenAI can change that cultural sentiment by making AVRs more accurate and less aggravating.
In the meantime, an AVR is essentially what DoorDash is doing here, albeit with a GPT-based brain transplant. Meanwhile, the blow to consumers may be softened by the fact that this will serve as more of a backstop for cases when restaurant personnel are too busy to answer calls.
Moving Target
So there you have it… just a few examples of AI being tested in the on-demand economy. And these aren’t the first. Instacart launched a ChatGPT-powered AI search tool known as “Ask Instacart.” Like some of the above attempts, this is all about recommendations and answers.
The remaining questions: Will we see others throughout the Localogy-verse do similar? Is conversational AI additive to food ordering… or is it a misaligned attempt to shove AI into everything? This will continue to be a moving target so we’ll keep our crosshairs locked in.
