Brief IA
›

Starbucks and ChatGPT: AI Complicates Coffee Ordering

Models & LLM·Tom Levy·

Starbucks and ChatGPT: AI Complicates Coffee Ordering

Starbucks and ChatGPT: AI Complicates Coffee Ordering
Key Takeaways
1The integration of ChatGPT into the Starbucks app complicates coffee ordering, making the process longer than with the traditional app.
2Users encounter conversation limits and location errors, hindering the effectiveness of the AI tool.
3The features offered by Starbucks, such as mood-based recommendations, seem impractical for most consumers.
Why it matters : The inefficiency of AI in simple tasks like ordering coffee raises questions about its real usefulness in everyday life.

A Complex Ordering Experience with ChatGPT

Ordering a coffee at Starbucks is a task that, on the surface, should pose no problem. Yet the recent integration of ChatGPT into their app has turned this everyday action into a real headache. For regular customers, such as a user who is used to ordering a venti iced coffee with light skim milk, the traditional Starbucks app makes it possible to place an order in a few quick clicks. However, using ChatGPT to accomplish the same task proved to be a path strewn with pitfalls.

Last week, this user decided to test the new feature for the first time. The process begins by opening ChatGPT and typing "@Starbucks" followed by the order. To his great surprise, instead of placing the order directly, ChatGPT began by providing a detailed description of the iced coffee, without ordering it. The interface then displayed a menu with several options, requiring the user to customize every aspect of the order, from the size to the type of milk, before it could be added to the cart.

Comparison with the Standard App

This process took far longer than the classic Starbucks app, where the user can simply select his usual drink and complete the order in a few seconds. Despite the complications, he persisted and added his wife’s drink, a fruity tea, to the order. ChatGPT suggested a lime iced tea, an incorrect suggestion, before the user remembered the right choice: the Passion Tango Tea. This extra step made the process even heavier.

Midway through the order, a worrying message appeared: "This conversation is approaching its limit." As a free ChatGPT user, he did not expect to hit this limit so quickly, especially for a task that was supposed to be simple. Moreover, ChatGPT recorded his location incorrectly, suggesting stores far from his current position. When he tried to correct this, he ran into an error message, and shortly afterward another message informed him that he had used up his available messages with the free model.

An Unconvincing User Experience

Despite these obstacles, the user decided to try again. However, the ChatGPT model he had been downgraded to could neither place the order directly nor even understand his earlier attempts. It simply guided him toward using the Starbucks app, making the AI integration pointless in this context. The experience highlighted the limits of the tool, which seems more like a technology demo than a genuinely practical service.

La lettre du soirGratuit · 19 h

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Starbucks seems to be aiming at an audience that does not really exist with this app. Their blog suggests orders based on mood or appearance, such as "Recommend a drink that matches the vibe of my outfit." These features, though entertaining, do not meet consumers’ practical needs. They may even make baristas’ work harder with complicated and whimsical orders.

The Dream of a Simplified Order

The idea of AI-simplified coffee ordering is appealing: an assistant that knows users’ preferences and places the order automatically. Companies like Google have tried this with voice assistants, and hope persists with projects like Gemini. However, ChatGPT’s current approach does not live up to this vision. Ordering a coffee should be a quick transaction, without requiring a lengthy conversation, especially before having the first coffee of the day. The current tool, far from simplifying users’ lives, seems to add an unnecessary layer of complexity to a task that should stay simple and direct.

The Limits of AI in Coffee Ordering

The integration of ChatGPT into the Starbucks app was supposed to revolutionize how drinks are ordered. However, it revealed significant flaws in using AI for simple tasks. The fact that the tool cannot efficiently handle a coffee order, a relatively basic task, raises questions about the effectiveness of AI in other areas of daily life. Users hope for a smooth, hitch-free experience, but often find themselves facing technical limitations that complicate the process instead of simplifying it.

A Technology Still in Development

It is clear that AI has a long way to go before it can effectively replace traditional ordering methods. Developers must take user feedback into account to improve these tools and make them more intuitive and practical. The ultimate goal is to create a user experience that is not only innovative but also genuinely useful and accessible to everyone. For now, the integration of ChatGPT into the Starbucks app remains a technological curiosity that still needs a lot of work to reach its full potential.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.