Computer-Augmented Latourex

10 min readOct 9, 2019

Today, visual storytelling wunderkammern The Pudding released Ye Olde Mad-Lib Pub Crawl Generator, by Caitlyn Ralph @earthto_caitlyn and Jan Diehm @jadiehm. It will plot a most-efficient path between any of 41 pubs in Great Britain that share the same name, e.g. “Red Lion” or “Pig & Whistle.”

This is the travel path for “Blue Bell.” This is only 10 pubs, since I’m such a lightweight.

Go check it out and indulge in the nerdery and the economic-linguistic history lesson and an introduction to the Traveling Salesperson Problem algorithm and the visual design. It’s lovely work, and exactly the kind of nerdery I’ve been collecting as part of my Semantic Noodling and Meaning Machines thread of inquiry.

This method of travel is part of a school of thought invented by Joël Henry and Rachael Antony called Latourex, for Laboratoire de Tourisme Expérimental. Joël’s twitter handle is appropriately @latourex. Latourex is a collection of “travel algorithms” or “games” that give travelers unusual, memetic (in the sense of affording retelling), and off-the-beaten-path ways to explore places.

While reflecting on the Mad-Lib Pub Crawler, I was reminded of the Latourex speculative design I created in 2012 at my prior gig as part of the series The Drawing Board. So here it is, reposted, below, with a few updates.

Nobody likes to feel like a tourist. When we look for guidance from typical sources, it can feel like we’re all working off the same script and we’re still not connecting to the real place. In this episode, we explore how people can use chance to find inspiration and authentic experiences when they travel.

Credits: Chris Noessel, pair-designed with Christina Worsing and Greg Schuler.

That video focuses on sharing the experience in a simple and cinematic way. But of course, the design rationale underneath is worth getting into in detail.

The Inspiration

Experimental Travel was developed by Rachael Antony and Joël Henry as a series of techniques to make travel more interesting. In this type of travel, people play “games” on the road, to get them off the beaten path. We love this idea where chance operations inspire and help support us in our search for unique and authentic travel experiences. In this Drawing Board, we explore how this approach might translate to a service design opportunity via a mobile experience.

The first thing we realized is that, as in Latourex itself, no single game would suffice. So we envisioned a container app called Latourex Travel Games that has an open and extensible architecture. Some game modules would be available at launch, but over time more modules could be added, keeping the content fresh. Lyrical Travel would just be one module within this container app, and the sketch below shows that the Latourex app is what Jeanette would download.

Once downloaded and launched, the app would briefly introduce itself, and then get Jeanette to the action as quickly as possible by providing a default selection from the available modules. If she wasn’t interested in this first randomly selected module, she could “roll the dice” (or shake the phone) to get another one. While fun and in the spirit of Latourex, that mechanism could become tiresome if she was looking for a particular module, so we provided means to select a particular one from a list if she wanted. Since we knew that such an app would grow by word-of-mouth, we also let her replicate the module through which she learned about the app: Lyrical Travel.

In the world of Latourex, Lyrical Tourism is a game where travelers select a song for a given location, and let the lyrics suggest things to do and see. When we imagined what this could be in a mobile app, we recognized we could keep the spirit of the original game and improve upon it by providing:

A huge database of songs that might fit given locations
Easy access to lyrics
Automated local suggestions based on those lyrics
Simple ways to capture, curate, and share the experience

We saw that we could facilitate this experience in four steps: getting started, guided suggestions, capturing key moments, and curating/sharing the results.

A Four-Step Process

1. Getting Started

In preparation for her adventure, Jeanette first tells the app some things about her travel: where she wants to explore; what her transportation options are; and how much time she has. A trusting traveler could even allow the app to look at their calendar or email to automate this step.

We suspect that most users would open the app just before wanting to explore, so the default screen uses her current location, allowing her to select another or even randomize. Rather than make these separate steps, we thought the map screen had enough real estate to let these be simple radio buttons or toggle buttons as overlays to the map. This way it would feel more like a single step even though it has three separate pieces of information.

We’re also big fans of the power of Lyrical Travel to encourage people to rediscover familiar places like their hometowns and suspect setting the current location as the default would encourage this idea when people open the app near home.

With a location set, now the traveler needs to select a song. By default, Lyrical Travel makes a suggestion, selected at random and based on a search of the location Jeanette has chosen. She has a control that lets her listen to it to see if she likes it. That makes it simple to use. We understand that a default song wouldn’t always be to her liking, so the app enables her to either “re-roll the dice” or allow her to select from her own music. (You’ll notice the interaction is the same from the prior screen. You can read more on the pattern, below.)

These two options — randomization or direct selection — work, but we also wanted to provide her an “in-between” option that would keep the interaction simple and still ensure an ideal match. To accomplish this, we thought that she could permit the app to scan her music collection (on the phone, or music service account like Spotify). In this scan it could look at any ratings she provided, combined with the sheer volume of songs she has of particular categories. From this, Latourex could infer the categories of music she prefers, similar to how Pandora can infer preferences from a small set of choices.

2. Guided Suggestions

With a place and a song in hand, the app next begins to make suggestions on what to do by looking for keywords in the lyrics. The algorithm would try to find more unusual words and unique phrases first, so the suggestions would be particular for each song, but gracefully degrade to more common words if those first ones don’t have any results. The app would then display the suggestions it found about local activities and place them alongside the lyrics, highlighting the connections if possible. Suggestions include an image, a title, and a small overview. With a tap, Jeanette can get to more details, including directions from her current location and the ability to rate suggestions. She is able to delete any suggestion with a swipe, giving room to other suggestions of greater interest.

3. Capturing Key Moments

Using persistent tools below the suggestions, Jeanette is able to capture moments through text, photos, and video. She’s free, of course, to follow these to the letter or interpret as she pleases. As she collects memories of the day, they stack below the suggestions, next to the lyrics. If she wishes, she can delete or rearrange these with standard iOS gestures.

4. Curating and Sharing the Experience

Jeanette is able to save and share her experiences through curated “video postcards.” We wire-framed three different ways for her to edit: Edit by ratings (the simplest), simple editing, and power editing. Below you’ll see the sketch for the Edit by Ratings screen, which lets Jeanette rate individual elements. The Postcard Player chooses and dynamically displays the elements during the duration of the lyrics, ensuring a minimum time per element, and augmenting elements with “The Ken Burns Effect” slow panning and zooming effects. Though we did not have time to comp it up, we thought that to be in the slightly-random spirit of Latourex, the postcard wouldn’t be a fixed video, but a semi-random display, like slideshows from a photo album app, that plays with the selected song and overlaid lyrics.

To share the postcard, Jeanette has the option to share with others by posting the video to the Latourex site, or sending a link directly to whomever she likes. There are some complexities about getting the media from her phone to the site, but we did not detail these out in this wireframe pass.

Design Challenge 1: Keeping Jeanette in the world

To meet Jeanette’s goals, her experience should involve interacting with technology but not being consumed by it. She shouldn’t walk around the city all day with the phone in front of her face. How do we create interactions that integrate into the flow of a person’s day but not interrupt it? To solve this problem, we made the interface as modeless as possible, letting her drop in and out of it as she needs or wants to.

Design Challenge 2: Avoiding musical burnout

If a single song serves as a guide, how is it offered in a way where the traveler doesn’t burn out on it by the end of the game? To solve this problem, we chose not to play the song by default. It was tempting to have the song underscore the lyrics as she was interacting with the lyrics, but it felt like it would be too much. It’s always opt-in, and the app would work well with her music player app.

Design Challenge 3: Varying degrees of control

When it comes to planning and travel itself, people want varying degrees of control. From highly supported and structured to free-form and interpretive, Lyrical Travel enables travelers to make decisions from direct selections to self-guided interpretation. This pattern of enabling people to move from simple to more complex decision-making processes is experienced across the service. For example, in the process of choosing songs people can accept suggestions; choose from random; or make their own, personal selections. In using the lyrics as a guide, people can select and follow specific suggestions in the order in which they appear in the song; jump around and choose at random; or disregard suggestions completely and interpret lyrics using their own internal compass. And finally, when creating a video postcard, as mentioned earlier, people can turn the reins over to Lyrical Travel or take full control of curating their day using the tools provided by the app.

Design Pattern: Random default first & Roll the dice

In several places throughout the app, we noticed a pattern emerging, and once we found it, we opted to stick to it. The pattern is that for choices Jeanette needed to make, we would not present a menu of options. Instead, we would first present a good, default selection. If she was in a hurry or wanted to keep it simple, Jeanette could just press OK and be on her way. And of course we would offer an option to select directly through a full-featured menu, but that didn’t feel complete given the spirit of randomness that’s inherent in Latourex. To meet that spirit, there’s a “roll the dice” option that would select another option randomly. And in the music selection, we even went one step further to take advantage of existing data to improve the likelihood of a roll Jeanette would really love. This pattern works especially well in the context of a semi-random travel recommendations app.

Service implications

In order to deliver on the promise of the ideas in the video, we realize that there would be not just an app, but a complete service underneath it, to help connect travelers and house the marketplace of modules. Imagining this container service led us to aspects of the design such as hearing about it through an existing member, letting similar travelers’ selections influence suggestions, and giving Jeanette options to meet other Latourex travelers near her.