TED Memory Project

We partnered with TED and Lilly to conceptualise, design and develop "The TED Memory Project":  a sensorial, participatory installation presented at TEDNext Atlanta, alongside a dedicated website open to a wider public. The project was designed to highlight Lilly's commitment to funding and advancing Alzheimer's research by placing real, human stories at the forefront, in a way that aligned with TED's human-centric vision. Our goal was to create an experience that reminds people just how meaningful and sensory our memories are, and in doing so, how important it is to care for them, because they are more fragile than we think.

From generating to capturing

At Domestic Data Streamers, we've explored the territory of memory in multiple projects (Synthetic Memories, Calling Home, and Forever Frequencies) often using generative AI to recreate or reinterpret past moments. This time, we wanted to take a different path. Rather than generating a personalised image for each visitor's memory, we designed an experience in which visitors selected multiple images from a curated gallery of 300 photographs to represent a memory of their own. The reasoning was deliberate: in the context of a live conference, generating a personalised image per person creates friction, as no image will ever live up to the expectation of an accurate memory. These images aren't meant to depict realistically a memory: they evoke situations, textures, and emotional contexts for each person to find a connection to their memory. It is the combination of the images each person selects, and the individual interpretation they bring to each one, that shapes their experience and makes it their own. Because the gallery is shared, the same image can hold entirely different memories for different people and without knowing it, strangers end up remembering different things through the same photograph.

These 300 images are the result of a hybrid workflow, combining human conceptualization and meticulous curation with generative AI, but their style was intentionally open-ended, blurry, slightly unfocused, sometimes caught in motion. They evoke a feeling rather than document a moment, echoing the warmth and grain of printed album photographs. To avoid bias in the selection, we built a specific framework to define the types of images we would generate, covering a diversity of situations, settings, genders, races, and locations, and tested the full gallery with a group of 15 people from different backgrounds and nationalities, refining the final set based on their feedback.

A space of reflection and connection

The installation consisted of three interconnected parts: 

1. Memory Booth for capturing your memory
2. Memory Gallery for discovering the collective album
3. TED Shorts panel for learning about the project and hearing some of TED's speakers share their memories

We intentionally designed the experience to hold two complementary spaces, one of individual introspection, and another one of collective dialogue.

In the Memory Booth, visitors were first asked a simple question: "What would you want more of in the future?" They could choose from seven emotions: connection, calmness, excitement, confidence, adventure, community, or security. From there, they were invited to think of a past memory linked to that chosen feeling, to select four images from the curated gallery that best represented it, and to write a short caption. Their album page was created on screen; they could save it to their phone, and choose to add it to the collective album. Before leaving the booth, they could see how their memory connected with others who had selected similar images, a simple reminder that their experience was part of something larger.

The Memory Gallery was a dynamic, evolving display across multiple screens, each one representing one of the seven feelings. As the conference unfolded, the gallery kept growing. Visitors could return at different moments to discover new memories and watch the collective portrait shift and expand. It became, in effect, a live data visualisation of what a community of people wants to feel more of.

Light, sound, and immersion

We wanted to create a pocket of stillness inside the noise of a conference, a space where visitors could slow down and genuinely connect with the act of remembering. A custom-designed suspended circular lamp covered the top of the Memory Booths, acting as both a visual anchor and a sensory layer that responded in real time to each visitor's interactions on the iPad. A sound layer accompanied the entire experience: from a voice-over introduction to a custom sound interpretation of each memory, built by mixing evocative sounds linked to the selected images: birdsong, waves, traffic, laughter. Above each screen in the Memory Gallery, a directional speaker emitted the sound of the memory being displayed, allowing visitors to connect with other people's recollections not only through images and text, but also through sound.

Beyond the conference

We adapted the on-site installation into a responsive website, making the experience accessible to anyone, not only those present at TED. The site kept the conversation alive, inviting users to contribute their own memories and grow the collective archive from wherever they were. This extended the project's reach beyond the conference, connecting it to a broader public conversation around Alzheimer's awareness and the importance of preserving memory.

Data Impact

What emerged from the memories collected was a portrait of collective longing.

• 135.000 web visits
• 381 memories created across the conference and the website
• 5,8 average minutes spent in the Memory Booth

What people want more of is connection (23%), followed by adventure (20%) and calmness (16%).

The most selected image was a plane, used in 84 memories, followed by a crowd and a tropical beach, each appearing in 48.

The words that appeared most frequently in memory captions were "sky" (118 times), "air" (94), "night" (93), and "smell" (73).

When asked to describe what they want to hold on to, people reach instinctively for the sensory, the open, the hard-to-name. The things we want most are, it turns out, the ones we can never quite pin down.