What does data taste like?

Coming up on Spark 227, contributor Alison Broverman visits the Bev Lab to taste a few tweet-based beverages. Here's a sneak peek:

Have you ever wondered what a tweet might taste like? No? Well Irwin Adam Eydelnant and Jarlath Byrne Rodgers have wondered about it a lot! The two met while pursuing PhDs in science at the University of Toronto, and quickly bonded over their passion for applying their science brains to food and drink experimentation.



About nine months ago they started a consultation company called I and J Ideations, which is devoted to looking at the future of food and drink in a somewhat off-the-wall scientific way.

For two weeks last month, Irwin and Jarlath opened up the BevLab, an open studio space in downtown Toronto where they invited the public to help them explore the taste of data. I went by to talk to them and check it all out.

Yes, one of their big experiments was the "data drink machine" - a project that allows them to explore the mechanics of tasting data.

Alongside a young computer programmer named Raul da Rosa, Jarlath and Irwin developed a program that pulls real-time tweets about food into a word cloud. Each of the food words in those tweets are coded to a certain flavour profile, which is reflected in an ingredient in the data drink machine.

On the day I visited, there were three ingredients - blueberry juice for sweet, lemon juice for sour, and ginger juice for spicy, standing in for salty - they want the drinks to taste good, after all!

The word cloud is connected to a microprocessor that controls a series of valves, and at any given moment, you can press a button to create a drink in which the flavours represent the tweets at that exact moment.


"What's fun about this project is that it's actually an integration of hardware and software," says Jarlath. "There's the conceptual side of figuring out how do we pull the data, how do we organize the data, but then how to we actually physically manifest that into a cup?"

But the question remains: Why try to taste data?

"We're dealing with massive data sets today, in our research and in our daily lives, and we have to have tools to understand how to interpret the data we're presented with." says Irwin. "There's a lot of visual strategies emerging, but often those visual strategies still remain limited in their ability to represent the data that they're showing... So the next question really can be how can we integrate more senses as well, that would be an obvious extension of this."

It all fits into the theme of BevLab. "We really are just using this as a space to explore beverage and food as something more than just a product that we consume and forget about," says Irwin. "So this is another theme on our message on food intent and food consciousness to start to think about what it is we are consuming, and have fun with it, of course."

And to build drink-mixing robots," adds Jarlath.


Comments on this story are moderated according to our Submission Guidelines. Comments are welcome while open. We reserve the right to close comments at any time.

Submission Policy

Note: The CBC does not necessarily endorse any of the views posted. By submitting your comments, you acknowledge that CBC has the right to reproduce, broadcast and publicize those comments or any part thereof in any manner whatsoever. Please note that comments are moderated and published according to our submission guidelines.