Sentient Scenes and Radically Adaptive Experiences
This video is featured in the AI and UX playlist and 2 more.
Summary
Sentient Scenes is a little toy with big lessons—an intelligent interface that changes style, mood, and behavior on demand. The project explores what it means to treat AI as a design material, weaving open-ended intelligence into interfaces beyond chat. We’ll share lessons in both technique and perspective for how the designer’s role evolves—from crafting visuals to sketching prompts—and what it takes to craft systems where machine intelligence mediates the user experience.
Key Insights
-
•
AI can be treated as a design material with unique strengths and weaknesses rather than just a productivity tool.
-
•
Radically adaptive experiences adapt style, mood, behavior, and layout dynamically in response to user context and intent.
-
•
Open-ended AI interfaces must operate within constraints to maintain cohesion and avoid 'robot fever dreams'.
-
•
Large language models excel more at parsing user intent than merely providing factual answers.
-
•
The prompt itself becomes a hybrid artifact acting as design spec, requirement doc, and programming language.
-
•
Designers’ roles evolve into defining the interaction sandbox and parameters for AI-user collaboration rather than specifying every UI detail.
-
•
Personality and presence in AI interfaces can be achieved without pretending to be human, using animation and interaction cues.
-
•
Bespoke UI assembles and rearranges interface elements in real time to fit user needs, reducing complexity in workflows.
-
•
AI can enhance accessibility by tailoring font, color, and interaction styles to users’ abilities on demand.
-
•
Sentient Scenes took less than a day to build due to using plain language prompts, highlighting opportunities for rapid, imaginative prototyping.
Notable Quotes
"AI is not just a tool for productivity or efficiency; it's a design material with unusual strengths and weaknesses."
"Sentient Scenes doesn’t just talk about the scene, it becomes the scene, adapting continuously to user input."
"We wanted to create personality in interfaces without pretending they are human."
"Radically adaptive experiences are conceived and compiled in real time, responding to context and intent."
"The big design activity here is sketching prompts, guiding how the AI should behave rather than drawing UI."
"The prompt becomes part design spec, part requirement doc, and part programming language."
"Large language models are far better at figuring out what you mean than simply giving an answer."
"With bespoke UI, you don’t present every single path to the user; you let the right tool appear immediately for the job."
"AI can transform accessibility by changing experiences to meet user needs in the moment."
"This isn’t about Skynet; it’s about simple signals of awareness that adapt to immediate context."
Or choose a question:
More Videos
"A big challenge is that a lot of governmental sites are filled with legally correct but totally incomprehensible content."
Yalenka Mariën Marie MervaillieDesigning for Digital Inclusion in the Belgian Government
December 8, 2021
"This veteran-centric approach gives us a foundation to grow outreach and better serve veterans struggling with PTSD over time."
Megan Nipe Lyndsay BoothHuman-Centered Design for Engagement: Maturing from Newsletterville to Personalized, One-to-One Messaging
December 8, 2021
"Local women researchers in crisis zones help ensure safety and allow authentic stories to emerge."
Dr. Jamika D. Burge Mansi GuptaAdvancing the Inclusion of Womxn in Research Practices
September 15, 2022
"Research teams need to be strategic about building tools to do ethical research, not just diverse."
Jemma Ahmed Robert Fabricant Sean McKay Llewyn Paine Kate Towsey Noah BondTheme Panel
March 11, 2025
"Rosenverse includes a fifteen thousand person strong Slack community dedicated to UX and product design."
Louis RosenfeldAbout the Rosenverse
March 15, 2026
"I fumed that my idea had been ignored, wondering if it was because I was new, a woman, brown, or had a PhD."
Nalini P. KotamrajuTwo Jobs in One: Being a “Leader who is a Researcher” and a “Researcher who is a Leader"
March 10, 2021
"An everything chat bot is almost certainly not how you’re going to build it; realistically you’re building three apps in a trench coat."
Sarah BarrettAI in Real Life: Using LLMs to Turbocharge Microsoft Learn
February 13, 2025
"Time and bandwidth are the eternal enemy of research."
Jane DavisStrategic Shifts and Innovations in User Research: Navigating Challenges and Opportunities
March 11, 2025
"Latino parents framed classrooms as the teacher’s domain and home as theirs, avoiding engagement to respect authority."
Alba VillamilStereotyped by Design: Pitfalls in Cross-Cultural User Research
March 30, 2020
