DoorDash is now accepting orders using simple language and images!
Ask DoorDash takes natural language or images of recipes and turns them into shopping carts. Here’s why conversational interfaces are a literacy issue for you.
On Thursday, DoorDash launched Ask DoorDash, an AI powered conversational interface that allows you to skip the tapping and scrolling and simply talk to the app to get what you need. Rather than navigating through different merchants to add items to your cart, you can simply tell the app what you’re looking for (“a filling dinner for 4 people” or “kid friendly vegetarian restaurants with mild food”) and it will create your order for you.
The image recognition piece is pretty cool too. You can take pictures of recipes from cookbooks, share your grocery lists, or send links to recipes and the app will parse out the ingredients and create a shopping cart for you, offering recommendations along the way. It will also remind you if you have something at home already (like sugar or butter), filter based on your dietary restrictions and price range, and tell you why it recommended a certain restaurant. Right now it supports restaurant discovery, groceries, and reservations (e.g. “a table for 2 in downtown at 8pm”). It’s available now on iOS in some areas but will roll out across the US over the next few weeks.
This isn’t the first time we’ve seen this kind of technology. In February, Uber Eats launched an AI powered Cart Assistant and Instacart has its own Shopping Assistant. But this is significant because once all of the major players come to the same conclusion, it becomes a trend rather than a feature.
Chat is quietly becoming the interface for everything
As someone who covers AI enabled development, why am I writing about a food delivery service? It’s because of the trend. We spent the last three decades learning ITS language (menus, filters, categories, checkboxes), but now the tables have turned. Rather than learning the language of the app, the app learns yours. You speak in sentences and the app performs complex tasks on your behalf. You take pictures of human created artifacts (recipes, grocery lists) and turn them into actions. Even though DoorDash doesn’t call it that, this is an agent.
As I always tell my students, if conversational interfaces are the future, then being able to accurately articulate your needs is the literacy of the decade. Someone who says “I want dinner” gets very different results than someone who says “I want a filling dinner for 4, one of whom is a vegetarian, and it should cost less than $60.” This holds true whether you’re using DoorDash, Claude, or any other service. When you place an order for food, you are training yourself in the most important skill of the decade. Don’t underestimate it.
Try Ask DoorDash when it reaches you — but treat it as prompt practice, not just dinner. Then steal the idea shape for yourself: photo of a messy thing in, structured result out. Open Claude Code and build a tiny version — 'read this photo of my handwritten list and turn it into a sorted shopping list with quantities.' One evening, and you'll understand this whole product category from the inside.
Major players are investing millions of dollars to normalize conversational interfaces with humans. As developers, this is a huge boon because by the time we get around to building our own AI tools, users will be trained to use them without any instruction. All they have to do is tell it what they want.
Source: techcrunch.com
Free course
Stop reading about AI — start building with it
The free Claude Code course: your first site, tool or game — no coding. No upsells, no cross-sells — nothing to buy here.
Start free →▌ Related guides

Author
Evgenii Arsentev
PhD · Chief Executive Officer, digital health
Articles · Latest articles