We all have that scrap of paper stuck to the fridge with a magnet. It has “milk, eggs, onions” scrawled on it in bad handwriting. The friction of moving those three words from the physical world into a digital cart is surprisingly high. You usually have to unlock your phone, open an app, search for each item individually, pick a brand, and add it. Uber Eats is trying to erase that friction entirely.
Key Takeaways
- Uber Eats launched the “Cart Assistant” AI feature in beta on Wednesday.
- The tool populates grocery carts from text lists or uploaded images of recipes.
- The assistant uses previous order data to prioritize familiar items and brands.
Uber Eats has released a new tool called “Cart Assistant.” It is a beta feature designed to do the heavy lifting of grocery shopping. Instead of searching for items one by one, you hand the AI a list, and it fills the cart for you. The goal is to speed up the process from deciding what to eat to actually paying for it.
The big deal
Online grocery shopping has a specific annoyance: the search bar. If you need twenty items, you have to perform twenty separate searches. That takes time and mental energy. By allowing you to upload a photo of a recipe or a handwritten note, the app removes the manual labor of building a cart. It turns a ten-minute task into a ten-second task.
This also signals a shift in how delivery apps compete. It is no longer just about who can drive the food to you the fastest. It is about who can make the ordering process feel less like work. Competitors like Instacart and DoorDash are already deploying similar AI tools, meaning this level of automation is quickly becoming the standard rather than a luxury feature.
How it works
You open the Uber Eats app, pick a grocery store, and tap a purple icon. You can type a list, or upload a photo of a handwritten note or a recipe screenshot. The AI reads the text in the image and finds the corresponding items in the store’s inventory.
Think of it like handing a shopping list to a clerk at a hardware store counter. You don’t walk the aisles yourself; you just hand over the paper, and they come back with a pile of parts that match what you asked for.
The software scans your input, identifies the products, and matches them to what is in stock. It also looks at your history. If you wrote “milk” and you always buy a specific brand of oat milk, it puts that specific carton in the cart. You can then review the cart and swap out brands if the AI guessed wrong.
The catch
The biggest potential issue here is accuracy. Handwriting is notoriously difficult for computers to read. If your “onions” looks like “lemons,” you are going to get the wrong produce. The system is in beta, which is usually code for “expect bugs.”
There is also the issue of brand selection. While the system tries to use your history, it has to guess when you buy something new. It might default to a sponsored product or a more expensive option rather than the cheapest one. You cannot just hit “order” blindly; you still have to audit the cart to ensure you aren’t overpaying or getting the wrong ingredients.
What now?
The feature is available now in the app for users who want to test it. If you are a regular Uber Eats user, try uploading a messy list and see if it actually works.
Watch to see if this actually saves time or if correcting the AI’s mistakes takes longer than just searching for the items yourself.