More than 7 million people each year came to the Boston.gov site to tap into all sorts of city services: reporting a pothole, learning the assessed value of their property, paying a parking ticket, checking recycling schedules.

Boston wanted to create a beautiful platform that would make the process of finding and navigating information easy and intuitive, one that prioritized the needs of its people.

The Boston.gov site had outgrown its existing structure and needed an overhaul. The user experience was mired by the fact that Boston City Hall has nearly 50 different departments, each with its own microsite, navigation and content standards that feed into a centralized site.

So the City of Boston hired IDEO to design its new digital home, Boston.gov. IDEO designers worked with the major stakeholders—department heads, employees, and Boston residents, with a focus on those in underserved communities and those with varying levels of technical proficiency—to gather insights into what users needed.

The objective was to design something that would simplify how people navigate the site, to make it easier for residents to find answers to common questions, and, in the process, serve as a community facilitator—an up-to-date, reliable resource that makes life in the city more manageable.

Rather than organize the site by the city’s departments, the design solution was rooted in illuminating relevant content by topics—people’s life moments—many of which overlapped through different branches of government. Owning a car in the city, for example, is closely related to preparing for winter, so bridging those common threads allowed for a more intuitive organizing principle.

In the process of the site redesign, IDEO also helped to create a visual identity system that underscores Boston’s strengths: the city’s confidence, humility, personality, and optimism.

In addition to the web architecture and redesign, IDEO also produced a comprehensive style guide to help maintain design consistency, so departments can add content without compromising the overall integrity of the site, as well as a brand look that the city has decided to adopt across nearly all of its consumer-facing media. Now you’ll see the famous underlined B on T-shirts, banners, and even snow shovels.

Holding true to its mission to include Boston residents and all site users in its design process, the new Boston.gov site was piloted for six months, and the process document on the city’s blog (including the IDEO designed brand guide), before a full launch was rolled out in July 2016.

“The website should act like a helpful human. This is one of the big differences between the old site and the new site,” said Jascha Franklin-Hodge, the city’s chief of information technology in a Boston Globe article. “Oftentimes, when you looked at something on the old site, it would feel like you were interacting with some sort of lawyer-robot that was speaking to you in government-speak, using very formal language. We are trying to move away from that so your experience is akin to a person.”

The City of Boston was in need of a new digital home—one that emphasized a shift to a more welcoming tone and commitment to its citizens.

‍

In the Western Hemisphere, The North Face brand has long represented an active, sporty lifestyle that’s filled with challenges and adventures. But appreciating “the great outdoors” in China is less about extreme sports and more about connecting with nature in a communal, spiritual way. This research-based insight led the team to identify new opportunities for The North Face to better target its messaging.

Designers found a key opportunity: Encourage Chinese consumers to escape the pressures of hectic city life by enjoying the freedom and renewal that comes with spending time outdoors.

Today, as people find less time to enjoy the benefits that lie just beyond China’s large cities, there’s a tangible desire to return to the outdoors to refresh, reassess, and renew oneself. The North Face aims to help consumers do exactly that.

To create a compelling new brand vision for The North Face in China, the team’s design research tapped into Chinese ways of spending time outdoors. IDEO’s insights and creative brief provided the essential brand story for The North Face in China and informed the tagline of the company’s current marketing campaign, “Go Wild” (去野). The North Face brand’s new messaging links “old nature” and “new outdoors” through a combination of digital, retail, and interactive community channels.

IDEO also collaborated with the company on designs and prototypes for a new digital platform and an updated store-within-a-store merchandising display. The digital community aims to attract newcomers and connect them with more experienced outdoors enthusiasts and clubs. As the platform evolves, it will help aspiring adventurers plan trips, learn skills, and share their experiences through photos.

The retail community space aims to integrate this digital community with in-store shoppers and an effort to turn the “buy-a-product moment” into a “join-the-outdoors moment.” Both will be integrated to provide a holistic brand experience in China centered on nature and community.

In the West, The North Face, with its decades-long history, has long symbolized an adventurous, active lifestyle. In China, however, the outdoor market only began accelerating in the early 21st century.

‍

Success in this growing market depends on building strong and deep connections with Chinese consumers.

‍

20% sales growth in China

from 2012 to 2013—the year the project was completed—propelling the brand from sixth place to a market-leading position

Over the last 36 years, IDEO and Lilly, along with the pharmaceutical giant’s other design, development, and manufacturing partners, have collaborated on more than 250 projects. During this time, Lilly has not only honed its engineering-focused internal device-development team but also transformed into a human-centered organization with extensive, robust in-house innovation capabilities.

Early in the relationship, a team of IDEO designers and Eli Lilly leaders traveled the world to understand how people with diabetes live with the disease. Their findings led to the design of a series of award-winning insulin injection pens like SAVVIO. Portable, aesthetically pleasing, and discreet, the pens are suitable for use in both medical and personal settings. Combined with Lilly’s technical design prowess, this human-centered focus allows the devices to seamlessly integrate into patients’ lives and routines, making a challenging disease like diabetes feel more manageable.

Fast forward to today, Lilly’s innovative, easy-to-use auto-injectors are not only continuing to provide life-saving treatments to people with diabetes but also enabling the adoption of the popular GLP-1 weight-loss drugs Mounjaro and Zepbound, which together accounted for the majority of the company’s annual revenue in 2025. Perhaps most importantly, patients consistently rate the experience of using Lilly’s auto-injectors as relatively pain-free, easy to use, and preferable to competitors’ brands, improving adherence, satisfaction, and health outcomes.

This unwavering focus on driving innovation through a patient-first mindset is evident in some of the most impactful projects Lilly and IDEO collaborated on over the years:

GLP Auto-injectors
Convenient and cutting-edge, the first Trulicity injection device was originally designed to deliver the GLP-1 drug to treat type 2 diabetes. Today, it has been adapted for use with other GLP-1s, Mounjaro and Zepbound, which are increasingly popular weight-loss drugs and are also administered once weekly via fixed-dose injections. A wide base enables the pen to be held securely against the injection site, and the device automatically regulates the dose and needle depth.

Taltz Injectors
For most patients, living with severe plaque psoriasis means being your own medical technician—metering your drug dosage each time, and attaching and removing needles by hand. Taltz provides patients with clear visual and auditory feedback throughout the injection process through simple icons, bright colors, a translucent material, and clicking sounds. The devices are available in both autoinjector and prefilled-syringe formats and are adapted for Emgality, a preventive monthly migraine treatment, and Ebglyss, which helps treat eczema or atopic dermatitis.

HumaPen SAVVIO Injection Pen
Key insights informed the design of SAVVIO, including patients’ preference for insulin pens that look less like medical devices and more like everyday objects you’d find in a backpack or purse. The sleek, colorful device was created to help patients feel comfortable and confident about fitting mealtime injections into their everyday lives.

Refillable Human Growth Hormone (HGH) Injector and Reconstitution Device
Designed to be kid-friendly, the Refillable HGH Injector leverages insights into pediatric treatment adherence to create a more integrated, systematic approach to treatment, while the Reconstitution Device makes mixing Lilly’s powder-and-liquid drug easier and less intimidating for parents and healthcare professionals alike.

KwikPen Insulin Pen
KwikPen is a next-generation, prefilled insulin injection pen that builds on the success of two previous designs, reducing its size to 5.7 inches and enhancing ease of use for patients.

HumaPen(R) Luxura™ Insulin Pen
HumaPen Luxura combines several innovations into a robust yet elegant pen that can be carried and used anywhere—an insulin pen solution that improves dosing accuracy, dose setting, and injection force.

In addition to designing products to meet the needs of people with diabetes, IDEO and Lilly have developed innovations in areas of oncology, autoimmunity, and men’s health. The results of this collaboration and expansion have won numerous design awards and produced one of the longest and most fruitful innovation relationships in IDEO’s history.

In the 1990s and early 2000s, Eli Lilly had a number of new, life-changing medications and two big challenges: Most of its drugs were self-injectables, not pills, so patients were hesitant to take them. And its drug-delivery device platform lagged behind its drug pipeline.

99%

of Trulicity users reported the auto-injector was “easy” to “very easy” to use in a JDST study

$36.5B

of Lilly’s $65.2B 2025 revenue came from sales of its auto-injectable GLP-1s, Mounjaro and Zepbound, a 45% increase from 2024

‍“…the Advisory Council will recommend transformational programmatic, cultural, and physical change to San Quentin that can serve as a practical model that can be replicated and scaled to other institutions.”

Source: Office of Governor Gavin Newsom: Governor Newsom Names Leading Criminal Justice and Public Safety Experts to San Quentin Transformation Advisory Council, May 05, 2023  https://www.gov.ca.gov/2023/05/05/san-quentin-transformation-advisory-council/

The Advisory Council included the expertise of established criminal justice, public safety, healthcare, academic, reentry, and rehabilitation professionals, along with system-impacted/formerly incarcerated community leaders, and representatives of crime victims/survivors. The premise of IDEO’s involvement is to augment the Advisory Council Co-Chairs’ and members’ expertise in evidence-based best practices and lived experience living and working in prisons, with expertise in collaborative design, synthesis, visualization, storytelling and bringing visions to life.

Specifically, IDEO was brought in to facilitate input-gathering, and assist with incorporating the many viewpoints into a single, cohesive report of recommendations to the Governor. This included meeting the challenge of bringing together the viewpoints of 21 people on the Council (17 members, three Co-Chairs and an advisor to the Governor); incorporating the views of people who are most impacted (those incarcerated at San Quentin, San Quentin staff, community-based organizations, victim/survivor groups, among others); and formatting and designing the final report.

IDEO’s facilitation included workshops with the full Advisory Council, individual interviews and check-ins with Council members to gather their input to inform the recommendations. Each Advisory Council member was asked to speak with, and record feedback from, no fewer than five people at San Quentin or in affected communities and to share this feedback with IDEO. IDEO integrated this information and feedback into this report.

Research activities included visits to San Quentin to speak with and facilitate conversations with:

• San Quentin residents (100+) including the Citizens/Inmate Advisory Council, focus groups with Spanish-speaking residents, former gang members, ACT (LGBTQIA+), Voices Heal, People in Blue, San Quentin Citizens Advisory Council.
• San Quentin staff (100+) including interviews with numerous officers.
• San Quentin staff and residents together.
• Current participants and staff at Amistad (MCRP) and Beacon (former lifer) campuses (many who were previously incarcerated at San Quentin).
  
  

IDEO also organized listening, information sharing and feedback sessions, both at San Quentin and in the community, between Co-Chair representatives and various stakeholder groups, including victim/survivor groups, labor unions, and advocacy organizations. These perspectives were also brought into the report.  

‍

Travel is fundamentally a human experience, and personalization has always been the ideal: the concierge who remembers your name, the host who anticipates your needs, the perfect room tailored to your preferences. 

But true personalization has traditionally been a luxury, reserved for those with wealth and elite status. Even as technology has enabled mass travel and big data and CRM systems have laid the foundation for personalization, brands have struggled to scale hyperpersonalization to everyone. Loyalty programs have become the industry’s best attempt at bridging this gap: Points and tiers quantify who deserves what, with status serving as a proxy for real relationships and customer understanding. 

Until now. With the introduction of AI agents, hyperpersonalization can scale from the segment level to the individual traveler. Done right, it has the potential to democratize what was once a luxury good: the experience of being recognized. 

Of course, serendipity and delight remain a huge part of the travel experience: unsolicited dinner recommendations from local shopkeepers, wrong turns that lead to hidden gems, and conversations with strangers that become core memories. Our research shows that, especially for Gen Z customers, agency and spontaneity play a big part in shaping their tastes and identities. The opportunity is for AI to know travelers well enough to make their experience feel personal—but also to know when to get out of their way.

Capturing this opportunity means redesigning loyalty so that rewards are legible to agents, personalization scales to the individual, and the system delivers an ecosystem of value matched to what each traveler actually cares about. In doing so, brands can finally deliver the kind of personalization that has always been the industry’s ideal.

The dual audiences challenge

Every travel brand is now designing for two distinct audiences. The first is the person who is comfortably waking up in their hotel bed, enjoying the aroma of morning coffee from the lobby brunch, or grabbing a beer with their new Scottish best friend by Fenway Park. The other is an AI agent that’s handling the searching, comparing, and booking tasks on that person’s behalf. 

These two audiences have different priorities, attention spans, and methods for evaluating information. Hyper-optimize a webpage for one audience, and you may inadvertently block traffic coming from the other.

It’s tempting to treat these dual audiences as a marketing problem—a matter of making pages readable to agents without alienating people. And that is part of it. Stripe has already built a wallet for agents, and AI coworkers are beginning to handle multi-step bookings. But the consumer trust gap is wider than the technology gap. The travel industry has not gathered enough data to determine whether travelers will allow an agent to optimize their preferences about brand, travel occasion, or loyalty points.

As brands strive to cater to both audiences, agents in the funnel are revealing three outdated assumptions baked into the way most travel and hospitality loyalty programs operate. First: transactions are the best measure of loyalty. Second: personalization doesn’t scale. Third: points are the primary reward. Once these assumptions are challenged, a different kind of reward emerges: the reward of travel itself. In the age of AI, loyalty programs should be designed with enough emotional intelligence to preserve the romance of travel while harnessing the superpowers of machine intelligence. 

False assumption 1: Transactions are the best measure of loyalty

When American Airlines launched its AAdvantage program in 1981, it introduced the first large-scale frequent-flyer program by mining its reservation system for recurring phone numbers. This marked the first time the airline could identify returning customers and begin building direct relationships with them rather than with their travel agents. The airline built the program on the data it had at the time: transactions. Forty-five years later, most loyalty programs still follow a similar structure—offering points for spending, tiers for hotel nights, and miles for distance traveled. 

This transactional dynamic is increasingly ineffective. The average American belongs to 17.4 loyalty programs, but actively engages in fewer than half. Many brands claim loyalty membership sizes that rival the entire adult population of the United States. These transactions may look like loyalty, but are they really?

The next generation of travelers will have an agent doing the work that only the most dedicated points gurus do today: discerning what each traveler finds valuable and surfacing the options that match. Imagine a Gold-tier member with a brutal travel schedule who is looking for a hotel close to tourist attractions. In a regular search, they filter by price, reviews, and location. An agent can do that, plus spot the perks that actually matter at six in the morning after a red-eye: early check-in and a quiet room away from the elevators. 

Now, the measure of loyalty broadens beyond the transactional into the experiential and emotional utility of being seen and cared for. Agents simultaneously reveal which programs offer real value and match them to what each traveler values for each and every trip. 

False assumption 2: Personalization doesn't scale

Top loyalty programs like Amazon Prime offer around 30 distinct benefits. They range from free next-day shipping to cash-back options and streaming media subscriptions. Brands surface long lists of these perks because they can’t possibly know which one will land with a given member. The list is often a coping mechanism for the brand’s inability to fully personalize and the consumer’s inability to remember everything the brand offers.

Marketers often use robust quantitative surveys, such as conjoint analyses, to gain insight into consumer preferences. Run enough permutations with enough travelers, and you can statistically estimate which combination of benefits they value most. Marketing teams still debate granular tactics, like the order of perks to list in an email. But those debates are entirely unnecessary when agents can just choose based on the traveler’s preferences.

Consider the scale this can reach. Expedia Group alone encompasses more than 3.5 million lodging and vacation rental properties. One can imagine that in the near future, by analyzing a traveler’s bookings and reviews across this vast selection, an agent can ensure that a guest’s preferences are prioritized for every booking, regardless of who operates the property.

Personalization scales when agents do the matching. The new design challenge is enabling agents to interpret hundreds of personal preference variations and reward ecosystems, making offerings sufficiently readable for an agent to match and communicate well on the consumer’s behalf.

False assumption 3: The points are the reward

Ask anyone who has been a road warrior what real loyalty feels like, and the answer is rarely the points. It’s often Shirley at the front desk, who remembers you like a higher floor, a firm pillow, and a bigger room rather than a view. The recognition itself was the reward, a relationship rather than a transaction. But for too long, it was too expensive to offer to anyone beyond elite travelers. 

Now, the behaviors that used to earn that recognition are fading just as technology is emerging to recognize customers without them. Today’s consumers are less loyal and less inclined to align themselves with a specific loyalty program or brand because they value choice, flexibility, simplicity, and convenience. The number of people who used to go on mileage status runs in December just to hit status is decreasing. Instead, they’re seeking something more, and brands are increasingly able to provide it, thanks to advancements in AI.

Expedia’s recent partnership with CLEAR offers a glimpse into the future of travel loyalty: It is evolving from individual brand perks to an integrated ecosystem of complementary benefits that enhance each step of the journey (think: Apple ecosystem logic, but applied to travel). For example, CLEAR offers Expedia members discounted memberships to CLEAR+ lanes and access to its Concierge services, providing VIP treatment as they head to their next destination. What stitches the ecosystem together is the underlying agent layer. It reads each traveler’s  preferences across the partner brands and surfaces the right benefit at the right moment, without the traveler having to coordinate it themselves. No longer just a perk for road warriors, recognition becomes an experience any traveler can enjoy at every stage of their journey.

The opportunity to make travel more human

People often talk about compound interest when it comes to money. You invest a small amount, it grows over time—that’s the magic of compound interest. But we believe the memories, stories, and friendships formed through travel compound even harder. Years later, a distinctive smell, a friend, a menu item—whatever—can magically transport you back to that travel moment. 

Loyalty has never really been about earning points. It’s about the memories of being cared for while away from home, discovering something new and unexpected, and feeling seen and recognized. It’s about the desire to return because those memories capture the joy of travel.

Nights, miles, and dollars are the receipts for a trip. They are not what the trip leaves behind, which is more sensorial and more personal than anything a tier structure has ever captured. The next era of travel loyalty will be about making travel more human, powered by the combined strengths of machine intelligence and emotional intelligence.

“In the age of AI, loyalty programs should be designed with enough emotional intelligence to preserve the romance of travel while harnessing the superpowers of machine intelligence."

I’ve never been a big fan of medical dramas, but I couldn’t stop watching The Pitt, the HBO Max show about a fictional Pittsburgh ER at a public teaching hospital. The high-stakes medical cases, relentless pace, and scrappy, sleep-deprived team you can’t help but root for had millions of viewers counting down to Thursday nights earlier this year. But for me, it was about watching the work itself. 

Apparently, I’m not alone. The Washington Post called the Emmy-winning show “perhaps the purest example of ‘competency porn’ on TV”—a term that surged into the cultural vocabulary this year to name the satisfaction of watching people who are really, really good at their jobs. 

But what they’re really good at goes beyond the practice of medicine. They excel at teaching and learning on the job. They think out loud, trust learners with calls that matter, and dissect what just happened before moving on to what’s next. I find it inspiring to watch.

Here are three lessons on teaching and learning in the workplace that we can take from The Pitt, and ways to activate them in your own work. 

Make mastery observable

Every scene on The Pitt has two things going on: someone doing the work, and someone learning to do the work. A recent nursing school grad shadows as the charge nurse guides a survivor through a rape kit exam—every choice a lesson in how to move at the patient’s pace. A resident observes as the attending decides out loud whether a procedure is worth trying if it might cost their patient her sight. A specialist performs an emergency procedure the attending has never seen before, teaching the room as she does it. Watching people at the top of their game work out a problem from beginning to end—including the messy middle, not just the outcome—is a sight to behold.

Most of us don’t work in a teaching hospital, but “see one, do one, teach one”—the century-old learning model attributed to Johns Hopkins co-founder and surgeon William Stewart Halsted—is how most of us pick up a craft, medical or otherwise. But that kind of learning is harder to come by these days. For years, most knowledge work has been evolving to a less readily observable state. Increased remote work makes it challenging to know what other people are working on, receive mentorship, get feedback from peers, or signal when you need help. The explosion of single-player AI tools has increased output, but made the thinking that informed it harder to see and, some research suggests, to trust. 

When my San Francisco colleague Thomas Overthun worked at Philips early in his career, he would pass by hundreds of drawing boards on his way out of the studio, displaying an amazing range of work in progress. Our workplaces look very different now, but it’s entirely possible to design tools, rituals, and ways of working that can help us see and get inspired by one another’s in-process work.

Here are a few practices that are helping us tilt our work toward each other:

• Use video shares: Kaii Tu in our Shanghai studio records short Loom videos to walk his team through the design choices he made—and why. The format forces him to make his reasoning legible, and makes it possible to share across time zones.
• Host “Screen Share Fridays”: On Fridays at our Cambridge studio, the studio-wide Slack channel explodes with screenshots of work in progress, sketches, project Post-it Notes, photography from the field, and presentation decks. Questions naturally follow, which then leads to dialogue about the work.
• Think in public: My Chicago colleague Leah Marcus builds incredible FigJam boards at the start of each project with relevant research, inspiration, and connections she’s making across our digital product portfolio. She uses plain language, bringing others who don’t share her business design discipline into her thinking.

Let the learner try

The best teachers I’ve observed share one hard-earned skill: they show restraint, letting learners grapple and find their feet with just the right amount of support. In The Pitt, when a new patient arrives in critical condition, the attending physician, Dr. Robby, turns to a resident and asks, “What’s your plan?” He doesn’t jump in, even when the clock is ticking. As a viewer, you feel the pressure that the resident is under and wonder, “Is he trusting a beginner with too much?” Turns out, this is an important part of the learning process.

This is the “do one” part of Halsted’s model, and it’s harder than it sounds. We know from decades of research that learning happens when we reach the edges of what we know how to do, and it happens even faster with scaffolding: a gradual release of responsibility from teacher to student. Instead of just telling residents to do the procedure on their own, Dr. Robby asks them questions, nudges them to consider alternatives, and fills in information they miss if it will have consequences for the patient. He lets them try first, then steps in when they get stuck. The scaffolding makes the risk survivable but not invisible. 

As a firstborn, Virgo, recovering perfectionist, I’ll be the first to admit that it is hard to use suggesting mode when you can just use editing mode. But the shortcut has real implications for my colleagues and the work. While I may get a cleaner draft, we both skip the interesting part that could help someone else grow, and probably get the draft to a better place than I could alone.

How might we demonstrate restraint and let learners try?

• Provide feedback in the margins: Suggest, don’t edit. Offer comments over rewrites. The extra friction is the point—it forces a conversation about the choice you would have made, instead of making it.
• Design it now: My colleague Bri Patawaran in our San Francisco studio is a huge fan of design-it-now moments to get a team out of swirl and uplift viewpoints that haven’t been heard. She asks, “What would you do if you had to deliver this today?” Everyone goes heads-down and then presents.
• Ask, don’t answer: When your team is stuck or a direct report brings you a challenge, resist the urge to solve it. Ask, “What's your first take?” and stay in question-asking mode as long as you can to help them explore their own thinking, giving you a chance to see their problem-solving skills in the process. 

Tell the story

The story we tell about what just happened is how we absorb what we’ve learned. On The Pitt, residents present each case to the attending after their patient is discharged. They walk through what the patient came in with, the diagnoses they considered and ruled out, the call they made, and the result. It’s a small ritual with an outsized effect—the resident consolidates what they learned by teaching it, and everyone within earshot learns alongside them. 

Reflection doesn’t have to be formal. Christopher Myers, a researcher at Johns Hopkins, studied medical transport teams—paramedics and nurses who fly in helicopters to accident scenes and rush patients back to the hospital. On any given day, they have no idea what they’ll encounter, and no single person can accumulate enough experience to be ready for all of it. What Myers found is that the crews built shared expertise through vicarious learning: telling each other stories, informally, between calls. These stories—casual, unscheduled, and recounted in whatever time they had—were how the whole team became more prepared together.

There are plenty of ways to build reflection into your work. The harder—and more interesting—work is building it into your culture. 

Here are a few practices we use across IDEO to encourage reflective storytelling.

• Ask “What are you working on?” The deceptively simple five-word question is a reflective invitation dressed up as small talk. Ask it in hallways, on Slack, at the start of one-on-ones. Half the time people just answer. The other half, they tell you something they didn’t realize they’d learned.
• Never skip a retro: It’s easy to nix a project post-mortem when you’re running short of time. But “I liked…,” “I learned…,” “I lacked…,” and “I longed for…,” is an easy and effective feedback framework to prompt everyone to share their ah-has and takeaways from the work. 
• Host a “Wrap Party”: Once a month, we host a virtual IDEO-wide “Wrap Party” where teams share three behind-the-scenes stories about recently completed work, giving everyone an opportunity to learn and get inspired together.

At one point while watching The Pitt, I said to my husband, “Wow, if I’d seen this when I was younger, I might have wanted to become a doctor.” (He gently reminded me that I have to avert my eyes every time I see blood.) Though what we see on screen may seem aspirational, none of the lessons The Pitt shares about teaching and learning on the job is exclusive to an emergency room. Observability, restraint, and storytelling are choices—small ones, mostly—that we can put into practice every day wherever we work. None of them require a shift change.

Curious about how to improve ways of working at your organization? Get in touch.

“The best teachers I’ve observed share one hard-earned skill: they show restraint, letting learners grapple and find their feet with just the right amount of support.”

‍

Imagine an engine that generates alternative endings for stories, a 2x2 visual tool for choosing which films to watch,  or a digital game of telephone that transforms a poem into something altogether different, like a location on a map. For Khyati Trehan, a Design Lead at Google Creative Lab, these are the weird, improbable, delightful explorations that AI is catalyzing.

Trehan’s lifelong curiosity for making began at Mirambika, a progressive school in New Delhi, the city where she was born and raised. “When you give children complete freedom, they choose to learn,” she says. That love of learning took her from the National Institute of Design to a globetrotting creative career spanning interning at a type foundry, creating AR experiences for Snapchat Spectacles and Instagram, producing 3D editorial illustrations for The New Yorker, The New York Times, and WIRED, working as a communication designer at IDEO’s former Munich studio, and “one crazy Oscars’ project the year of Will Smith and Chris Rock.” 

She’s now at Google Creative Lab in New York, working on projects that “humanize complex technology and remind people why they love Google.” She collaborates with a diverse team of designers, writers, and technologists, united by a desire for creative freedom and a novel approach to problem-solving.

‍I spoke to Trehan recently about how AI enhances her creativity, her evolving mindset as a designer who embraces coding, and the importance of fostering play and experimentation in organizations to promote AI adoption.

‍Ed White (EW): Do you remember when you first started using AI?‍

Khyati Trehan (KT): I got early access to DALL-E before it launched in 2021. Ironically, I remember liking the blurry loading states more than the final images. The results never quite stuck because nothing that came from my prompts felt like…me. There wasn’t much control over the output. I was looking for tools that would feed my creative process, and “one-shot AI”  didn’t do it for me.

An interest in using LLMs to turn natural language into software code happened more recently. My friend Pedro Sanches, a brilliant creative technologist, designer, and Creative Lab alumnus, came over for tea and shared some sketches he’d made using what he called “coding sans coding.” (This was before “vibe coding” was a term.) I was much more excited by that approach because it felt like something I could incorporate into my practice by building tools to help me explore new places without them feeling completely unfamiliar.

Those early experiments provided initial insights into my personal philosophy on how and when I use AI. My goal has never been to rely on AI to do all the work; instead, it’s about exploring how AI can help me supercharge and advance existing ideas.

EW: How is AI changing your team and Google as an organization?

KT: As the boundaries around our disciplines blur, the number of hybrid creatives is growing. Writers are making films, graphic designers are building writing tools, animators are engaging in creative coding, and developers are designing interfaces. There’s still a distinct difference between the vision and “flavor” of what I might design and develop as a graphic designer and what a creative technologist might create using the new capabilities that AI unlocks for both of us. Our individual experiences, values, core strengths, skills, and knowledge still define what makes us unique, even when we have the same tools at our disposal. 

EW: What are the things you and your team are learning, as designers, about using this technology?

KT: I think we’re realizing that regardless of how you look at the creative process, the shape of it remains the same. We still start by playing and experimenting to explore the edges of technology. We engineer every aspect of the applications we create, ensuring that we consider people’s needs. In fact, we now spend even more time and energy focusing on what matters to people and asking ourselves: “When we can make anything, what do we choose to make?”

While engineers are trained to focus on efficiency and optimization—and form the backbone of Google—creatives, when given the opportunity to lead research and gain early insights, instinctively seek out emotion and play. We naturally discover the right metaphors and interfaces that make complex systems clear.

EW: What’s worrying you about AI and design, and why?

KT: With every big shift, it’s wise to be cautiously optimistic. I often reflect on how we studied design in school: manually painting a color wheel and using our judgment to create the right shade of orange that would sit between red and yellow, even when the Blend Tool existed in Illustrator. Putting time and effort into these exercises sharpened our skills and helped us develop a strong foundation, which remains useful regardless of the tools we use. We shouldn’t forget that. You can use AI as a crutch, or you can use it to unlock or supercharge your existing skills, expertise, and ideas. To me, that’s the distinction between an effective use of AI and slop.

EW: What are the really concrete ways AI has changed your craft as a designer at Google? 

KT: The tools you use change the way you think. When I was learning 3D modeling and texturing years ago, it quite literally unlocked a new dimension in my graphic design practice. I’d surprise myself with the ideas that came to me and what I was capable of with this new ability.

This holds true for using AI. I’ve added surfaces like Gemini Canvas, AI Studio, and Flow to my toolkit to choreograph Google’s models and APIs. Mindset-wise, it’s made a lot of us hybrids. Now that we can build things, I find that in meetings, designers show more often than they tell.

I still design using traditional tools, except now, once I translate the visual and the flow into a clickable prototype, I can iterate in the same environment, and the work becomes more lived-in. 

EW: How has that changed you as a creative, and why?

KT: It’s definitely a leap. It’s like the difference between learning about qualitative interviews versus being in the room yourself. You can understand both objectively, but with the latter, you feel more connected to the learnings. For example, I’ve designed loading states plenty of times, but now that I’m closer to the front end, I’m bringing so much more of my design flavor and delight to them. I feel more comfortable taking risks and taking departures from what’s deemed standard in the space. 

EW: What’s an example of that?

KT: I’ve been designing digital experiences for very personal, everyday, specific, and idiosyncratic needs. For example, I got my hands on Ted Chiang’s Exhalation, a collection of short stories, and fell in love with them. I’ve been inventing alternate endings for some of them and extending his beautiful worldbuilding.

This led to the creation of Story Arc Engine, a narrative-building tool that allows users to deconstruct stories using a five-part narrative arc. By tweaking one part of the arc, users can see how a change in the plot affects the rest of the story and generate new narratives based on their own plot ideas.

EW: How else does AI change what you’re designing? 

KT: Because I’m now building both the final output and the intermediary tools that help me get there, I share both: the final product and the tool I vibe-coded to create it. This means that others can use the tools I make, often in unexpected ways. For example, someone once used Story Arc Engine to draft their next career move, hiding a sabbatical in the narrative. 

EW: What other examples of these types of projects have you been working on?

KT: I’ve recently made several tools for myself. Around the World in Good News is a digital newspaper that explores uplifting historical events and stories of human achievement from across the globe and throughout time. Another project, 2x2 Anything, emerged from my desire to make more informed decisions for movie night. It's a concept-mapping experiment where you define two conceptual axes, set the context, and click anywhere on the coordinate map to generate a fitting concept or summon an existing result. My most recent sketch, Machine Telephone, is a playful game in which you enter an input, pass it sequentially through different media and models, and observe how context shifts, translates, or gets misunderstood over time. For example, you might see a poem translated to a specific location on the map or a song transformed into a spherical material. Now that mediums feel less siloed, and those that remain siloed are easier to learn about, my first instinct is no longer to dismiss an idea just because it initially seems unfeasible.

EW: What advice would you give leaders who want their orgs to use AI effectively?

KT: An urgent tone from execs and leadership, coupled with a top-down mandate and a lack of concrete guidance, isn’t helpful. It only leads to confusion and stress, and, ironically, slows down the actual work

Instead, explicitly give people permission to play. Play is a powerful tool, especially when you’re faced with ambiguity. Give people access to a variety of tools and functions, and offer learning resources. Let people stumble upon new paths, and figure out which tedious parts of their process AI can take off their plates. 

EW: How do you think AI will transform your industry over the next five years?

KT: AI is already transforming industries by narrowing the gap between different disciplines. When the mechanics of creation are no longer a bottleneck, and the effort it takes to make things decreases, where does our time and attention go? It comes down to the core of why we create: our taste, the sum total of our experiences, our irrationality, our perspective, and our vision. Maybe we’ll just learn more about what makes us uniquely human along the way.

“Design in the Age of AI” is a series of conversations with designers and makers from across industries and disciplines, building the future with AI, today.

“My goal has never been to rely on AI to do all the work; instead, it’s about exploring how AI can help me supercharge and advance existing ideas."

When we prototype, we embrace what we call the three Rs: rough, rapid, and right. We keep prototypes rough, so that the people we design with feel comfortable providing honest feedback. We move rapidly, getting in as many reps as we can to open up creative possibilities, learn from each iteration, and avoid falling in love with the first ideas that pop into our heads. And we work very hard to get it right, creating lo-fi designs that test for just one variable at a time, so that, for example, a beautiful design doesn’t distract from a less-than-ideal experience. As AI makes it easier to generate polished renders with minimal effort, we see more people celebrating the ease and speed of design. But beautiful images aren’t design. In fact, they can seduce us into thinking an idea is fully baked, stopping us from asking the most important questions, the ones that uncover the insights that help us truly meet human needs. 

For decades, the three Rs have kept us grounded, reminding us why we prototype and ensuring that we get the most out of the design process. They apply to everything we design—from physical products to AI-enabled systems and digital tools. One of our most memorable prototypes was for a digital product for Sesame Workshop. It involved a simple cardboard cutout of an iPhone and a bit of role-play. One colleague pretended to be a preschooler, standing in front of the board, pointing to elements on the improvised screen. Another stood behind the board, impersonating a furry monster that introduced a new dance move whenever our “preschooler” tapped the invisible screen. It did exactly what a prototype is meant to do: It drew people in, sparked a discussion with our partners at Sesame Workshop, invited valuable feedback, and could be easily adjusted and improved, giving us numerous opportunities to make it into something kids and parents loved. The video we recorded became so popular that professors still show it in design schools today.

As we integrate AI tools into our prototyping process, it’s crucial that we use them thoughtfully and intentionally, so we don’t lose the methods that have enabled us to create category-defining innovations that meet real human needs. We have to maintain the three Rs to expose knowledge gaps, invite real feedback, build belief, and give us confidence that a concept is worth developing and investing in. They help us make space for surprising turns, illuminate opportunities for joy in the design process, and even let us embody and fully engage with our ideas. Here’s a deeper dive into why each R matters, and how to integrate them while embracing AI tools.

Rough: A strategic incompleteness

Early prototypes aren’t finished or even pretty. That’s on purpose—it gives the people we test with permission to dive in and co-create with us. It also keeps us from becoming overly attached to our initial ideas, making it easier to pivot if we see a better way forward. 

When our toy inventors created the first prototype of the Aerobie Rocket Football, for example, it featured a small pedestal glued to the bottom of a foam football. The goal was to create a football that could stand upright, allowing kids to practice place-kicking without someone holding the ball. The pedestal was clunky and awkward, so we replaced it with fins. Nothing fancy—just four fins cut from a sheet of foam then hot-glued to the ball. It was better, but still incomplete. The next iteration included a ball with two curved parting lines running from the nose to the tail. Taking advantage of this feature, we attached the fins along the ball’s curve, creating a helix. The design compromised stability, and the ball immediately toppled when we tried to stand it up. But when we picked up the ball and threw it, it flew through the air with a perfect spiral. We had accidentally made something better than what we were aiming for—a toy so successful that it’s been selling for decades.

How can you intentionally introduce roughness into your prototypes when your tools automatically default to high fidelity? One approach IDEOers use is to roll them back to earlier versions, such as Midjourney 1.0, where figures are rendered with multiple fingers and other noticeable flaws that indicate a work in progress. You can also ask AI to create sketch-level or monochrome outputs rather than full-color rendered ones, leaving things visibly incomplete, annotate prototypes with open questions, or present multiple competing directions side by side, ensuring that no single option feels like the answer. The goal is to signal to collaborators and users that their input is still welcome and that the design has room to evolve.

Rapid: More cycles, questions, and opportunities to learn

We don’t want to spend too much time on a single round of prototypes. The goal is to iterate as quickly as possible, giving us multiple opportunities to share our ideas with users and to grow and morph our designs into something that resonates functionally and emotionally. By addressing challenges and responding to feedback, we reduce risk and build confidence with each iteration.

For example, when we prototyped a new voting booth for Los Angeles County, we wanted to incorporate tactile and audio interfaces for individuals with visual impairments. Instead of fully coding the experience and pre-recording audio clips to respond to testers’ actions, we hired a voice actor and improviser. He was hidden in an adjacent room with one of our designers, who had taped a flowchart to the wall. When our “voter” tapped the keypad to navigate the voting process, the designer would point to a phrase on the flowchart, prompting the voice actor/improviser to read it into the microphone. The audio was then relayed to the voter via headphones. Throughout the research session, our designer could make last-minute adjustments and explore different paths on the flowchart, enabling him to quickly test which scenarios provided the best experience for our voters. The flexibility of the prototype—combined with the voice actor’s improvisational skills—allowed us to modify not only the flow of the interaction, but also the wording, pacing, tone of voice, and more, on the fly.

There’s no doubt that AI accelerates output, but that’s only valuable if the time saved goes back into more cycles, more questions, and more variations. How do you make sure AI’s speed translates into more iterations rather than locking in your design too early? Show your prototype to various stakeholders to solicit their feedback. Listen closely and incorporate what you’re hearing. If you’re working on a digital design, for example, and the people you’re testing with don’t like the size or placement of a button, or how an interaction works, consider how you might alter the UI. Does moving the button over a tad solve the problem, or is there something else that isn’t working? With AI, you can make changes and test them almost immediately, even if you don’t have a coding background. However, be sure to keep your updates focused on the feedback you’re hearing rather than on testing a completely new design. 

Right: Testing one variable at a time

“Right” means building a prototype early in the process to answer a specific question, rather than showcasing the entire vision. When too many variables are included in one prototype, it can be hard to discern what is working and what is not. Do people love it because the ergonomic shape feels great in their hand or because they like the color or texture you chose? Or do they dislike it because of how it functions mechanically or digitally? Splitting these attributes apart allows you to learn about each element separately before bringing them all together into a cohesive prototype in the later stages of refinement.

An example of this is Flipslide, an electronic game that we designed and licensed to Moose Games. The toy was inspired by the fidget craze, which includes items like fidget spinners, chew beads, pop-its, and the resurgence of the Rubik’s Cube. Recognizing this trend and wanting to create a challenging gameplay experience without a screen, we came up with a game concept in which pieces flip, slide, and click into place as players manipulate them to match tiles that light up and change color. We built our first prototype to test the movement of our design and understand what it would take to create a mechanical toy with shifting components. The initial model was made of foam core and rubber bands. Next, we wanted to create a fidget toy that felt satisfying to hold and transform, so we made a prototype with more bulk, incorporating machined plastic parts held together with tension springs that provided just the right amount of resistance when moving the pieces around. By refining this model, we created a prototype that felt particularly satisfying to hold and manipulate. It also made a fantastic clicking sound as pieces fell into place. 

After we solved the mechanical elements, we moved on to gameplay. We started by taping a stack of 2 x 2-inch laminated color grids onto the front. Each grid had a different color pattern, representing rounds of gameplay. To complete a round, rectangular wings had to be rotated, flipped, and moved into place to match the colors on the grid. Once the colors matched, we’d yell “ding!” and pull off a grid to represent the next round, initiating another flurry of rotating, flipping, and sliding to match the new color scheme. Satisfied that the gameplay felt fulfilling, we added electronics to bring sound effects, music, pacing, and more advanced gameplay into the test. Soon after Flipslide hit the market, we had an award-winning game and fans around the world.

Don’t get caught "boiling the ocean”—using AI to address every question in one prototype. Instead, build discrete prototypes for discrete questions. Before generating anything, ask, “What am I trying to learn? Who needs to weigh in?” AI is exceptional at producing artifacts, but you need to direct it to give you artifacts that teach you something. You want to create prototypes that you can bring into research and put in front of users to understand what resonates with them and why. That’s easier to do when you know what you’re testing for.

Maintaining joy and discovery

It’s easy to use AI tools to manage a creative process. Some are even designed to do parts of it for us. But as designers, it’s crucial to invest in our own learning by doing. When we get our hands dirty, we learn to recognize what’s right, engage our imaginations and deep thinking, and energize ourselves and our teams. That practice is how we achieve great outcomes, for ourselves and our clients. It’s also how we keep it human—putting ourselves in the shoes of the people who will live with these designs every day and experiencing them ourselves.

There’s no question that AI is a game-changer for designers, and one that we’re happy to embrace. But it’s crucial to use these tools to support the best of design, rather than replacing the methods that have led to some of the world’s most important innovations. Building prototypes that enable you to co-design with users, test assumptions, and iterate with new learning leads to better results. Experiment with different tools, including the latest that AI has to offer. Once you understand each tool’s strengths and limitations, you’ll better understand what combination you need to engage not only others, but also yourself. Because ultimately, prototypes and great design will always need human creativity, imagination, and judgment. And the three Rs are a great way to get there.

“For decades, the three Rs have kept us grounded, reminding us why we prototype and ensuring that we get the most out of the design process.”

I lead IDEO’s long-term strategy and emerging growth initiatives.

My career spans entrepreneurship, financial services, and technology, anchored in designing sustainable businesses that make systems and products work better for people. I have led multi-partner innovation portfolios with organizations including Citibank, Fidelity, Nasdaq, and the Gates Foundation, focused on building more human-centered futures. 

I began my career as an entrepreneur, creating new ventures in healthcare and real estate. I hold an MBA from Harvard Business School and a BA with distinction from Bates College.

I lead IDEO’s business globally.

A longtime IDEOer and a champion of creative collaboration, I bring a strong belief in the power of design and creativity to solve the world's most complex challenges.

Before taking on the role of CEO, I helped grow IDEO’s global presence, co-founded the Tokyo studio, and launched D4V, Japan’s first early-stage design-focused venture fund. I became a Partner in 2017 and was appointed to the board of IDEO.org in 2020, a role I continue to hold.

While I was away from IDEO, I served as Chief Creative Officer at Moon Creative Lab, a venture studio that powers the creation of new businesses for Mitsui & Co. There, I guided the development of new ventures and led Moon Media, a storytelling initiative exploring creativity and entrepreneurship through film.

As a committed educator, I have taught at esteemed institutions including Copenhagen Institute of Interaction Design (CIID), New York University’s Robert F. Wagner Graduate School of Public Service, and The University of Tokyo. I hold a degree in cognitive neuroscience from the University of California, Berkeley.

‍

Over the course of my 45-year IDEO career, which originally started by meeting David Kelley at Stanford in the mid-1970s,I have worked as a design engineer, a project leader, a business relationship leader, a studio leader, and a practice leader. I have helped us build and nurture many key long-term client relationships, including Silicon Valley tech firms, Fortune 100 consumer businesses, and health-care companies of all sizes. I have been leading or guiding the “Design for Health” side of IDEO’s business for more than 20 years. I have also been named on 55 granted patents.

I am an Adjunct Assistant Professor at Stanford University’s Hasso Plattner Institute of Design, known as the d.school, where I have taught at least one course each year for the past 45 years. I have led courses on product design, engineering design, and human factors design, as well as design for sustainability, and creativity and innovation. I currently teach the “Design for Healthy Behaviors” course with Dr. Nancy Cuan, M.D. and IDEOer YC Sun

I have a BS in Mechanical Engineering with an emphasis on Industrial Design from the University of Notre Dame and an MS in Product Design from Stanford.

I founded IDEO because I wanted to work in a creative environment surrounded by people I admire. These days, that’s still the case. I get to work with inspired people who have a commitment to making great things happen in the world.

I also founded Stanford University’s Hasso Plattner Institute of Design, informally known as the d.school. As Stanford’s Donald W. Whittier Professor in Mechanical Engineering, I am the Academic Director of both of the degree-granting undergraduate and graduate programs in Design within the School of Engineering, where I have been a student or a teacher for nearly 50 years.

I am most passionate about using design to help unlock creative confidence in everyone from students to business executives.

Together, my brother and I co-authored the New York Times best-selling book Creative Confidence: Unleashing the Creative Potential Within Us All.

Most AI products today are built for work—a space with clear problems and established systems. One founder noticed a gap: personal life is messier, harder to systematize, and mostly left behind by the AI boom. So she built Cue, a personal AI that lives inside iMessage and group chats, to go where the other tools haven't.

‍

In this episode, Becca Carroll, IDEO's Chief Strategy Officer, talks with Sida Li, co-founder and CEO of Shared Context Lab, about staying anchored to a human need while the technology around her keeps changing shape, why she treats her business model with the same rigor she'd bring to a product, and what it feels like to build a company at this particular, disorienting moment in AI.

‍

The conversation also gets into how Sida makes design decisions: the language Cue uses to describe itself, the tradeoffs behind building inside iMessage instead of a new app, and a real story about a business idea that didn’t pan out.

‍

This is the second in a two-part series profiling founders from IDEO's Startups-in-Residence program. The first conversation is with Johannes Seemann, founder of Sooner, on designing GenAI for the emotional side of money.

Most personal financial tools are built to run the numbers and optimize towards a budget. While that works for some, most people experience money as a lived relationship that does not neatly fit into a spreadsheet. Johannes Seemann and Becca Carroll discuss why money is emotional before it is mathematical, and what a human-centered approach to building a generative AI financial product looks like in practice.

Mina Seetharaman talks with Scott Shigeoka, author of Seek and Head of Curiosity Cultivation at the Eames Institute, about what distinguishes genuinely curious leadership from performative curiosity, how power dynamics shape curiosity, and why practicing curiosity can restore energy rather than drain it.

Mina Seetharaman talks with David Epstein, author of Inside the Box and Range, about why total freedom often produces average work, how leaders can design useful limits on purpose, and what organizations such as General Magic and Pixar reveal about the relationship between boundaries and creativity.