The New Design Canvas in the Age of Intelligence - Part 1 Relationships
In Brief:
The 3R Framework (Relationships, Roles, Reach) captures how AI transforms what we design - this series explores each R, starting with relationships
Our digital interactions are evolving from transactions to continuous learning relationships
This shift unfolds in three ways: from discrete sessions to continuity, from generic to personal, and from static to adaptive experiences
For designers: we're now architecting evolving relationships that require careful consideration of trust, memory, and boundaries
What are we designing in the age of AI?
What are we really designing in the age of intelligence?
This question hit me a few weeks ago as I was designing a profoundly generative AI-assisted tool for customer service.
Just three years ago, we only designed well-defined, rule-based, graphical UI flows that users navigated in mostly isolated interactions. Now, working on several AI-heavy design projects, I’ve felt we need to redefine what we’re designing.
Let’s explore this transition in a three-part essay series called The New Design Canvas - The 3R Framework. The framework, undoubtedly a work in progress, aims to highlight the core areas of design that are shifting as we move into the age of intelligence.
The new design canvas for 2025 and beyond focuses on three R’s:
1. Relationships - How our interaction with digital experiences is going from isolated interactions to deeply personal relationships (this week)
2. Roles - How we as users are going from operators to directors (next week)
3. Reach - How digital experiences are going from isolated islands to interconnected ecosystems of AI agents, other software and people (in two weeks)
For part 1, let’s unpack how our interaction with digital services evolves from isolated experiences to deeply personal relationships.
1. Relationships - from one-off interactions to evolving relationships
Consider the digital services you currently use in your daily life. If you’re like me, most of them rely on impersonal one-off interactions - from buying clothes to purchasing a metro ticket and checking the weather.
In the coming years, we’ll design increasingly personal AI assistants to help us tackle our professional and personal goals.
This means three shifts in our relationships with digital services as users.
1. From Sessions to Continuity
Instead of discrete, independent interactions with digital services, we’ll increasingly engage in continuous, evolving relationships with our bit-based helpers.
The services will learn how to best serve our needs and tastes through these evolving interactions. For instance, Duolingo’s language learning app’s new Roleplay feature adapts its teaching style based on your learning patterns and progress.
Duolingo adapts to users as they progress on their learning journey.
For designers, this evolution to continuous relationships brings about a need to consider questions like:
How does the relationship start? How do we make it feel friendly, helpful, but not obtrusive?
How does the relationship grow deeper? How does the service get to know more about the user in a way that adds value safely and securely?
How does the system surface past memory in relevant contexts?
2. From Generic to Personal
How well does your digital workout tool or energy company app understand you?
As the broad trend of AI-driven experiences scales, we’ll increasingly use services that understand us deeply. They evolve from generic to profoundly personal.
Many of our engagements with digital services will include deep context about our past actions, preferences, tastes, customer relationships, etc.
Instead of simple recommendations - you might like this Nordic noir thriller because you watched a similar one before - we’ll increasingly use services that form a comprehensive understanding of our mental models.
Cleo, an AI financial assistant, evolves its guidance based on your spending patterns and financial behavior. It analyzes your transactions and adapts its communication style - from weekly expense reviews to savings recommendations - to match what motivates you best. Through its chat interface, it builds a deeper understanding of your financial habits over time, making its advice increasingly personalized rather than just offering static budgeting tips.
Cleo, an AI financial assistant, adapts its tone and advice based on your spending habits - notice how it maintains a conversational, friendly style while delivering key financial insights and gentle nudges about spending.
As services know more about us, we as designers need to ponder questions like:
Beyond legal compliance, how do we build trust with users through transparency, security, and tactful explicit data collection?
What is the tone and intimacy of the relationship? What are the boundaries of the interactions?
How does the service reference sources for its information? How transparent should it be about its learning data?
3. From Static to Adaptive
A third layer of the transformation in our relationships with digital services is their shift from static to adaptive.
Digital services we use adapt and evolve based on our shared history with them.
This is already apparent in our use of AI assistants like Claude and ChatGPT. I have a months-long business strategy thread with Claude that has evolved and adapted to the uniqueness of my business context. As my thinking about my business and career has evolved, so has our conversation.
GitHub Copilot provides context-aware code suggestions based on your current work, using large language models trained on public code. It offers adaptation through custom instructions and organization-specific models that can align with your team's coding practices and conventions.
GitHub Copilot offers real-time, context-aware code suggestions, adapting to your current project context while allowing teams to customize its behavior through specific instructions and custom models.
Could these examples foreshadow how digital services co-evolve with us in the future?
Instead of a standard cooking app, your kitchen coach evolves with you as your tastes, skills, and health change. As the relationship deepens and our user experience co-evolves with the service, it can also deepen our customer relationship with the company. We’re less likely to switch to a competitor when the service fits like a glove.
Soon, the gap between our most useful and intimate AI-based apps and traditional generic and static services will feel jarring.
What does this shift in relationships mean for design?
When designing AI-enabled experiences, we're no longer just creating interfaces - we're architecting evolving relationships that deepen over time.
Most of the questions raised here are entirely new. We’re starting to design dynamic and deeply personal relationships with intelligent assistants, not just simple machine learning-based recommendations from yesteryears.
This means designing for progressive trust-building, thoughtful surfacing of shared context, and clear boundaries that respect the intimate yet artificial nature of these relationships.
The key is to make the system's growing understanding visible and valuable to users while maintaining appropriate distance and user control.
Thinking about these new relationships is a fascinating challenge on its own. Next week, we’ll stir the pot even further with a deep dive into how our roles as users evolve from operators to directors.
- Matias
