Eyeo Festival Day 1

Here are my very rushed notes from some of the talks on day 1 of the Eyeo Festival in Minneapolis, where “artists, designers and coders build and bend technology and give us a glimpse into what’s possible, into what’s next. Ones and zeros float all around us just waiting to deliver the next new interaction. The Eyeo Festival brings together the most intriguing and exciting people in these arenas today.”

I’m enjoying learning to navigate the elevated system of tunnels (the Skyway) in the city, which seem to get as much use in the summer as they do in what I understand are some cold winter months. The Walker Arts centre is a great venue, and I keep on finding new corners that the festival has spread into. I feel a little sorry for the regular museum visitors, surrounded by designers and coders, when all they want is some quite time with Edward Hopper.

See also: Workshop | Day 2 | Day 3


Mike Bostock – Visualizing Algorithms

Mike Bostock

Mike Bostock is a visualization luminary working at the New York Times and author of D3.js, a Javascript library that a lot of people use for putting together complex, performant visualizations. His talk focused on the use of data visualizations as a way of better understanding and testing the inner workings of an algorithm.

He used a set of classic algorithms for sampling,  sorting, shuffling and maze creation as a way of showing the value and pitfalls of the use of visualizations, how they can both better express what’s going on in the machine while also potentially misleading if not carefully crafted. A geeky talk that somehow felt like it had broader meaning when it comes to understanding the value of showing data visually.

His argument here was that visualizing the inner working of algorithms was useful for a number of reasons including teaching code, debugging and simply entertaining.

P1000147 (1900x1257)

Most of the visualizations were highly dynamic but insightful, although Mike emphasized that often a lot of work is required to have an algorithm output data and state in a way that can be used visually. The following image shows a dynamic visualization of a Fisher-Yates Shuffle, which he compared to a number of other shuffling approaches. The lines start out ordered and fanned out, and gradually become more messy as they swap places.

P1000148 (1900x1069)

In the following image Mike showed a long but static image that showed how items changed place during a QuickSort.  The vertical strips gradually line up in colour order the nearer you get to the bottom.

P1000149 (1900x969)

In the following static version of the earlier image, Mike showed the inner workings of a Mergesort. It shows how arrays of sorted numbers gradually merge to a solution.

P1000150 (1900x1021)

Mike showed a number of dynamic visualizations of algorithms for maze creation which demonstrated the diverse ways that each generated a pattern procedurally. It was hard to see the paths through them, though. Most effective was a moment when he took one of the visualizations below and peeled off the route to show it as a tree diagram. That, for me, was a cool moment of showing off.

P1000151 (1900x1025)

P1000152 (1900x1036)



Giorgia Lupi – The Shape of my Thoughts

Giorgia Lupi

Giorgia  works at Accurat, producing static and interactive data visualizations that showcase a whole range of topics. She talked about the profess of sketching, of collecting inspirational images, often from the world of art, and of appropriating them into her work.

Much of her appropriation is done through Pinterest. It sounds, like many  designers, she is an inveterate hoarder. You can see her Pinterest boards here.

A very blurry shot of Giorgia…

P1000161 (1900x1382)

Some sources of inspiration (top of image) and the visualizations they inspired, many of which you can see on Acurat’s site:

P1000155 (1900x1006)

A sketch and intermediate visualization for the Lifecycle of Ideas, featured in Popular Science, which shows the rise and fall of citations in a number of scientific fields.

P1000156 (1900x1056)

Sketches and images for History of 100 Geniuses of Literature.

P1000157 (1900x1017)

Examples of appropriating and reconfiguring images from modern art.

P1000158 (1900x981)

Sketches of Giorgia’s boyfriend. I like the fact that these aren’t great (he thinks they look like George Lucas) but she persists.

P1000160 (1900x1012)



Sarah Williams – DigitalMatatus: Visualizing Informality

Sarah Williams

Sarah works at MIT, where she runs the Big Data Design Lab and is involved with various aspects of urban design. Her talk started broadly, looking at the history of data design, and ended with a detailed look at work done to visualize the transit system of Nairobi, which is based on a self-organizing network of “Matatus” – small minibuses that seat 8.

Really enjoyed the pragmatism of this talk, the engagement with the different constituents in Nairobi, the clear development of a data set and visualizations of the 130 routes in the city, and finally the “appropriation” by the government of the resulting transit map.

P1000165 (1900x1337)

Sarah started by doing a short historical tour of big data, going back through Roman trail maps, the impact of the printing press, and the redlining maps from the 50s (example from Baltimore below).

P1000168 (1900x1069)

Sarah’s elements of civic data design:

P1000169 (1900x1326)

A guy who makes number for a Matatu bus once the driver has a license to allow him to drive on a route.

P1000170 (1900x1534)

The digital map of the Nairobi bus system, generated using GPS data and mobile phone fieldwork. The map has been adopted officially by the city.

P1000172 (1900x1069)


Brian House – Quotidian Rhythms + Political Frictions

Brian House

Brian is a artist who works with sound and data. He covered a broad range of funny and critical pieces of work that he’d done in an area that, for me, can seem a little trivial – translating data into music. I came away impressed with some of his work while also feeling that a few pieces were based on one-liners that didn’t sustain my interest.

P1000173 (1900x1379)

The Quotidian Record – 11 minutes of music representing one year of personal tracking data. Unusually for this kind of project (of which there are many) it is both quite musical, and beautifully presented.

P1000174 (1900x1279)

P1000175 (1900x1127)

You’ll Have to Take My Word for It – a piece for two guitars and saxaphone, based on the black box data from the crashed car of the Lieutenant Governor of Massachusetts. The data contradicts his account of the crash in many way.

P1000176 (1900x1602)



Cesar Hidalgo – The Bit and the Atom

Cesar Hidalgo

Cesar Hidalgo runs the Macro Connections group at MIT Media Lab. His talk centred on issues of information, our attitudes towards communication and the limits imposed on us by the instantiations of data in physical form.

I enjoyed his talk although it was all a bit tidy and fluid, and I was hit by a wave a jetlag that meant I DID nod off a few times. He’s a dynamic speaker, so it wasn’t really his fault.

P1000180 (1900x956)

P1000181 (1900x1132)

Stefanie Posavec @ Eyeo
Eyeo Festival Day 2