Keep In Touch
Timely, personalized suggestions to help you keep in touch with people.
Project summary
Goal: Help people keep in touch with friends and family by helping them recognize and take advantages of opportunities to connect about everyday happenings.
Solution: Timely, personalized suggestions based on apps people use every day.
Status: Prototyped
My role: Concept, design, prototyping
AI can help you keep in touch
This project shows how AI could help you keep in touch through timely, personalized nudges that work with apps you use every day.
Keeping in touch, in practical terms, means reaching out to someone about something that connects you. You could have just read an article, watched a show, or went to a museum that you thought they would appreciate. Perhaps you saw something on the street that made you think of them. Or maybe a memory materialized of the last time you hung out. Maybe your phone even reminded you of it, as a iOS Photos or Facebook “memory”.
This doesn’t necessarily take much time, but it takes effort. Keeping in touch requires an awareness of whom to reach out to, about what, and in what medium. In the constant noise of modern life, which pulls our attention in many directions at once, it can be difficult to have that awareness. Many of us want to keep in touch more with friends of family, but it can be a challenge to take action.
This project had 3 goals:
Explore how AI can help us keep in touch with people we care about.
Explore how technology can help people get more value from their data (another project, Selftalk, is an example of this).
Practice designing and prototyping using native iOS components straight from the source: Apple’s iOS UI kits for Figma. UI kits let designers work faster by giving them building blocks to get started, and ensuring that designs reflect what is feasible to implement.
Now let’s make it more concrete.
The following examples illustrate how timely, personalized AI suggestions built right into your phone could help you keep in touch.
All app screens shown, except Google Maps, were recreated in Figma by me– not screenshots (view in Figma).
Restaurant recs, and following up on them
If you like to go out to eat, you’ve probably gotten a restaurant recommendation. Trying a new restaurant can be a simple way to connect over a shared experience. It’s a quick excuse to share something about your trip while showing the other person that you appreciate their recommendation.
Sharing a recipe
It’s easy to connect over food and cooking. Sharing a recipe can be a great reason to reach out to someone. Maybe your friend mentioned that they’re looking for new recipes, and asked if you can send them some. While this is a good opportunity to keep in touch, it can be easy for this small request to get buried under other to-do’s. AI remembers so you don’t have to, and prompts you to take action while it’s top-of-mind.
Complaining about the weather
Talking about the weather might not be the most interesting way to reach out to someone. But even the most anti-weather-talker has found themselves complaining about the heat or the rain when they need something to say. For some, the simple act of checking the weather can give you a reason to reach out to someone.
Going to a concert
If you’re like me, sharing music with friends is one of the best ways to keep in touch. It’s a nice feeling when someone ends up liking an artist that you shared, or even better, goes to their concert. Sharing that experience with your friend can be an excellent way to connect.
Trash talk
We’ve looked at some nice, considerate ways to reach out to friends and family. But if good-natured ribbing is more your style, there’s always trash talk. If you’re a sports fan, you might already be familiar with this one.
Finding a doppelgänger
Have you ever found someone’s doppelgänger in the wild? It’s like finding a glitch in the matrix. Whomever you’re reaching out to will want to hear about it.
Design notes
Now, I’ll break down how this project came together and how it was inspired by the upcoming launch of Apple Intelligence.
Selecting UI components
To create these examples, I used the “shortcut” and “tip” components from Apple’s iOS UI kits for Figma.
I used the “shortcut” component because this feels like an extension of the existing iOS “Shortcuts” feature, which lets users set up automated triggers for actions that they take within apps.
These are basically auto-created notifications, so I aimed for the same design language that Apple is already using.
The examples show 2 types of notifications:
1. A notification about something specific onscreen. This uses the “tip” component which calls out the relevant area onscreen.
2. A notification about everything onscreen. This uses a variation of the Shortcuts “parameter confirmation” component.
A visual language for AI
The goal was to make the AI to appear lightweight and fully integrated with other apps. It should be clear when you’re getting an “intelligent” suggestion, and also what prompted it.
Apple’s upcoming Siri redesign essentially makes Siri appear more integrated with the entire phone. I wouldn’t be surprised if, in the long term, they drop the “Siri” name entirely, absorbing it into the iPhone brand itself.
I didn’t see the Siri redesign until after doing this project (I swear), and was pleasantly surprised to see are a bit similar.
Apple Intelligence: what to look for
I can’t wait to see what Apple will do, in the long run, with Apple Intelligence. User-specific context is the key ingredient for building the best AI products, and no company has more of it than Apple. Because Apple has an iPhone-sized portal into the lives of the 1.5 billion richest consumers on the planet, they are better positioned than any other company to create AI products that truly know their users.
The definition of “AI” has always changed, and will continue to change. For now, Apple defines “system intelligence” as:
How the system works with the apps people use every day to make “the everyday” easier for people.
Making products feel like they know you and understand you.
AI’s been there the whole time
iPhones already use AI in subtle ways that many users barely notice. For example, when your friend sends you an address, your iPhone automatically copies it to Apple Maps. Or have you noticed that your phone suggests apps for you to open based on the time and your location?
Small things in the background, barely noticeable, can save you time. What if all the things worked like that?
The launch announcements to date have mostly been focused on creation features: tools for writing and creating images. That’s fun, but it’s not where the biggest value lies. Eventually, I think AI-enabled iPhones will help save people time and effort on tasks like calling customer service, finding parking, or booking a haircut.
Is Siri back?
The most interesting part of Apple’s public announcements around Apple Intelligence so far, for me, has been Siri. Yes, Siri, the virtual assistant maligned by many, including Larry David. Siri could be the poster child for tech companies overpromising and underdelivering. Simply put, it’s just not good enough today to deliver on expectations that Apple has created.
But combined with AI, Siri is a major opportunity. For many iOS users, Siri can be a good introduction to the concept of one’s device as a virtual assistant that improves their day-to-day lives by augmenting their abilities.
When everything is AI, nothing is.
When everything is AI, nothing is.
AI is just a way in which the technology that we use gets smarter. It’s not really a “feature”, it’s something more integrated with the whole system. In the long term, as “AI-powered” becomes the norm and users’ expectations reset, maybe the “AI” buzzword will go the way of appending “.com” to company names: it will just be assumed.
Short-term, the reality is that companies face immense pressure to capture customers and shareholders attention through hype, marketing, and buzzwords. In a market that rewards hype, where mentions of “AI” on company earnings calls really are correlated with stock performance, it’s more or less rational for CEOs to say into every product announcement.
Laughing, but serious
In this project, I went straight past the simplest (and perhaps more relatable) use cases in favor of more novel ones.
But the capabilities shown here could save users serious time on everyday tasks like finding parking, signing up for a workout class, or ordering takeout. These are tasks that people spend significant time on, so solutions that save time can make a big difference.
The goal of building products hasn’t changed.
AI will raise the bar for product-builders to understand what’s valuable to their users, so that they can personalize products to better deliver that value. But the goal for designers remains simple, although difficult in practice: learn how users really use your product, where they struggle, and what they value. Then, go solve for those.
All app screens shown, except Google Maps, were recreated from scratch in Figma by me and are not screenshots (view in Figma).