Machine Scans Twitter For Mentions Of Fruit, Then Turns Them Into Smoothies

Which would you prefer? A boring old pie graph, or strawberry-banana data deliciousness?

We tend to think of data in one way: numbers. We can change the fonts, or we can graph them out. Maybe we can even use advanced visualization to make someone really grasp a particular figure. But numbers are always using one sense: sight. Could meaningful data be consumed in other ways?


Quite literally, yes it can. Tasty Tweets was a project by CIID students Kat Zorina, Ruben van der Vleuten, and Kostantinos Frantzis that condensed mentions of certain fruits across Twitter and mixed them into smoothie blends. Over five days of experimentation, 10 different fruits made their way into smoothies with the same relative proportion as their appearance in tweets. So if people talked more about pineapple than strawberries, then smoothies would taste more like pineapple. The idea is data you can taste–what the team calls “a graph in a glass.”

“The goal was to break free and think beyond traditional means of data representation and analysis,” the design trio tells Co.Design. “The graphical approach to data visualization has been explored for a long time and we hope that we can at least trigger others to explore the possibilities of visualizing data by different means or senses.”

Interestingly enough, the 1:1 representation of fruit as flavor wasn’t the project’s original intent. At first, the team was actually attempting to create liquids of certain colors to coordinate with Twitter trends. They’d manipulate the pH of a solution to effectively changing the color of a chemical compound within the liquid. “Although we achieved this, the distance between the message and the user seemed too broad,” the team explains. “Therefore, we developed a more basic concept that incorporated additional dimensions and senses.”

Their literal representation of data is equal parts elegant and brilliant. But the great thing about Tasty Tweets is that it could actually scale beyond fruit mentions. Imagine if, when you woke up each morning, there was a smoothie waiting that summed up the day’s news. You could know if people were talking about different regions of the globe (by their regional fruits) or what the big topics of the moment were (religion could have one flavor and the economy another). This concoction might not be all that specific, but packing big trend information into something we’re going to consume? Masking information within our favorite flavors? Why not?

[Hat tip: FlowingData]

About the author

Mark Wilson is a senior writer at Fast Company. He started, a simple way to give back every day.