The Funny Maps | 9 Humor-Computer Interaction Lessons Learned from a Sassy Google Maps
Part 3 of “The Funny Maps” Series — by Hannah Mernyk and Brooke Sachs
So here’s the fun part, testing actual sassy turn-by-turn voice navigation with actual people in the actual outdoors.
Before we could start the pilot, we needed a route and a script. We plotted a 10 minute out-and-back walking loop and brainstormed possible sassy and funny comments that our imagined navigator could make along the way.
Some gems below:
We knew when designing the commentary for testing, we wanted to include both context-specific sass and more generalized sass to see if there would be any difference in people’s responses. We baselined this by posting a survey with a 1–5 Likert scale of “Not Funny at All” to “Hilarious”online. We picked the highest ranked ones to record, with the exception of the Benny F. comment above which was mixed but which we had a suspicion would do better in person. We recruited a sassy friend to record the ten we selected.
Fifteen hours, ten volunteers, and two protocols later we can share some of our insights from the Funny Maps experiment. Spoiler alert: Things didn’t go as planned.
1. Sometimes king-size candy isn’t enough.
Because we are honest human beings, we told our volunteers they would need at least 20 minutes to complete our pilot (10 for walking, 10 for pre and post questions). It was hard to pull folks off the chairs for this length of time, even though we promised candy and told them it was for ~science~. It also didn’t help that it was intermittently snowing/sleeting/raining/freezing in late April.
2. Technology is fickle.
We couldn’t embed our recorded comments in the actual Google Maps UI because the turn by turn directions were never consistent — plus we wanted to simulate the sense of “I am using this nav because I don’t know where I’m going” so we Wizard of Oz’ed this by having people wear wireless headphones synced via Bluetooth to a tester’s phone, and the tester would play the comment at the appropriate time. Sound complicated? It was. To make matters worse, neither of us had our own Bluetooth headset. Fully 5 of 5 scheduled volunteers had to be sent home because the Bluetooth didn’t sync or because they didn’t bring a Bluetooth headset with them. A lesson to aspiring physical prototypers: BRING YOUR OWN TECH. Repeat: BRING YOUR OWN TECH.
3. Creativity is super hard to measure.
We tried using the Remote Associations Test (“RAT”), which is a word association task, to gather information about people’s level of creativity before and after our protocol. Even though it has been used as a creativity measure in other psychological reviews, it straight up didn’t work for us. Figuring out how to measure creativity is a whole other can of worms.
4. No, YOU hurry up….
Teasing is a specific kind of humor that occurs among friends and family members, not usually an AI. We hypothesized that your sassy navigator friend could act this way, so we played a “Hurry up, Gramps” midway through the walk. This was not well-received — people thought it was either a command, or just plain rude (insert Michelle from Full House gif here). It could be the case that with repeated use and more familiarity with the digital agent, people may accept a friendly barb or two, but our exploratory study indicated that this was the upper limit of sass they could tolerate.
5. People can’t estimate 200 feet to save their life.
Self-explanatory. Typical driving navigations offer distance-based markers but this was not at all aligned with a walker’s mental model. Users suggested using landmarks or a time marker (“turn right in a minute”).
6. Oh sorry, I guess you had to be there.
The jokes that our participants enjoyed the most were:
- One that referenced the reality of stressful student life at Carnegie Mellon (“look at these trees — barely hanging on, just like CMU students”)
- Two jokes making fun of items in their immediate environment (see art and Benjamin Franklin images)
Our other sarcastic and lightly needling jokes weren’t nearly as popular as the ones above. The success of the CMU joke isn’t all that surprising, since psychology tells us that people love inside jokes that not everyone will relate to, but the other finding is enlightening. Creating more jokes referencing environmental features could not only be key for making Google Maps funny, but may be helpful in making all digital agents more successfully humorous. In addition, people not only found these references funny, but also voiced their feeling that the navigation was getting them to look at their surroundings more mindfully. This bonus advantage of mindfulness might be able to make users more zen, as well as make them laugh.
7. What this doesn’t work for:
Based on user feedback, we heard loud and clear that people would not want to use this style of navigation audio when time was a factor, like if they were running late or were on a strict schedule. They also said that it might be too distracting during driving, especially if it points out features in their surroundings (no one wants to explain a fender bender by admitting they were looking for Benjamin Franklin in shorts).
8. What this could be great for:
Given that a humorous navigation audio should be reserved for walking over other means of transport, users said that they would enjoy using this when navigating in a new place or city. Others said that they might want to turn it on as a way of making their boring, routine walk to work more interesting and entertaining. Finally, some of our participants mentioned that especially with the environmental references, it might be great to use as an audio museum guide or for college campus tours. This way, by keeping the user in a confined area and taking out the sometimes stressful element of wayfinding, listeners could focus on and enjoy the commentary more.
Throughout this piloting process we learned a great deal about humor in a semi-controlled context with a specific audience, but there’s a lot more to explore. Due to our MacGyvered testing system, it’s hard to say how people’s reactions would change if they could see the typical visual interface… would they be more relaxed with a visual fallback and be more open to using a sassy audio? Would they actually use it in the situations that they described to us? Also, we’ll have to figure out a whole new way to measure creativity if we want to gather information about the supplementary advantages to humor aside from entertainment. Seeing as this was just a pilot study though, we gained a lot of insights to help inform how a formal experiment on this subject should be conducted.
Incorporating humor into technology is the next frontier for digital agents. We are moving beyond automation to “augmentation,” where humans and machines work together to complete tasks or make decisions. This leads to more complex conversations and longer interactions with the agent. Increased likability and error forgiveness are already proven uses for humor in tech, but that’s just the tip of the iceberg — facilitating creativity, exploration, and deeper engagement with the physical world are not-too-distant possibilities. Our pilot is just one small step for (embodied digital wo-)man, and one giant leap for humor and HCI.
