Designing with AI

Gemini 2.5 Flash Image editing: a game-changer for rapid storyboarding

From Ars Technica:

"AI image editing allows you to modify images with a prompt rather than mucking around in Photoshop. Google first provided editing capabilities in Gemini earlier this year, and the model was more than competent out of the gate. But like all generative systems, the non-deterministic nature meant that elements of the image would often change in unpredictable ways. Google says nano banana (technically Gemini 2.5 Flash Image) has unrivaled consistency across edits—it can actually remember the details instead of rolling the dice every time you make a change."

There are all sorts of fun examples online that demonstrate how well Gemini 2.5 Flash remembers details of the source image.

From a UX perspective, my favorite use cases for Gemini 2.5 Flash are storyboarding and persona representations. Imagine uploading a photo or illustration of your persona and prompting, "Show the user looking overwhelmed while multiple notifications pile up," and then "show them calmly focused with notifications smartly organized in the background"—all while maintaining the persona’s exact appearance across frames. Gemini 2.5 Flash keeps consistency across images far better than past models or my previous experience with dedicated AI storyboarding tools.

To get started:

1. Go to Gemini.

2. Make sure you select “2.5 Flash” (top left).

3. Click on the “tool” icon & “Create images.”

Google improves Gemini AI image editing with “nano banana” model
⚡ Quick Read (3 minutes)

Designing for AI

From user problems to AI solutions: a practical framework

In a recent issue of his newsletter, Design with AI, Xinran Ma shares his approach to strategically integrating AI into design thinking. Instead of starting with "what can AI do?", Xinran advocates for a bottom-up approach that grounds AI adoption in actual user problems. He provides a helpful, concrete example of this approach: the customer experience of obtaining a driver's license in Vancouver, B.C.

The first steps involve having a clear understanding of user problems and then mapping a relevant AI capability that can help solve that problem. This generates a preliminary set of AI-powered ideas that can then be evaluated and prioritized as a team. With AI, a key prioritization factor is the level of risk: with low-risk/high-impact ideas having the greatest potential. Finally, you can use AI to help you develop your high-potential ideas further.

I love how clear and concise Xinran's explanation of the process is, with practical examples and prompts.

How to strategically integrate AI into design thinking
☕ Medium Read (6 minutes)

AI Product Development

How and why your product playbook needs to change for AI

If you've been wondering why AI products sometimes feel unpredictable, Gian Segato explains, in a recent essay, the fundamental difference between the deterministic software that has dominated the past 30 years and the probabilistic nature of AI:

"⁠We're moving away from deterministic mechanisms, a world of perfect information and perfect knowledge, and walking into one made of emergent unknown behaviors...Building probabilistic software is like nothing we've done before...Our products can now succeed in ways we’ve never even imagined, and fail in ways we never intended."

Some key takeaways:

• User expectations need recalibration: people expect deterministic outcomes (press button → get result), but AI delivers probabilistic ones (press button → get one of many possible good results).

• Testing and validation methods must evolve: traditional A/B testing and success metrics don't capture AI's emergent behaviors.

• Embrace experimentation over perfection: the goal isn't 100% reliability but finding the right balance between control and capability.

While this piece gets technical in places, the core insights about how AI fundamentally changes product development are essential reading for anyone involved in building AI-powered experiences.

Building AI products in the probabilistic era
🔍 Long Read (17 minutes)

Broader Implications of AI

Mass Intelligence: AI for everyone, everywhere

Ethan Mollick, in an essay this past week, argues that with the broad adoption of more powerful AI models, we are entering an era of Mass Intelligence:

"So here we are. Powerful AI is cheap enough to give away, easy enough that you don't need a manual, and capable enough to outperform humans at a range of intellectual tasks. A flood of opportunities and problems are about to show up in classrooms, courtrooms, and boardrooms around the world. The Mass Intelligence era is what happens when you give a billion people access to an unprecedented set of tools and see what they do with it. We are about to find out what that is like."

This essay is an excellent reflection on the current state of AI and an important read for considering how broadly AI is becoming an integral part of everyday life. (Also - per Gemini 2.5 Image Generator, he's got a fun example of its power involving Neil Armstrong and the other Apollo 11 astronauts.)

Mass Intelligence
☕ Medium Read (8 minutes)

That’s it for this week.

Thanks for reading, and see you next Wednesday with more curated AI/UX news and insights. 👋

All the best, Heidi