Sharing the desk with a robot 🤖

Notes from the talk we held at Blend 2017.

ILLO

Published in

ILLO Stories

6 min readJul 4, 2017

3/3

Hey, we’re https://illo.tv/ a design and animation studio based in beautiful Turin, Italy.

For those who were there in Vancouver, and for those who weren’t fast enough to get a ticket (as the event sold out in just 12 hours!) we’re sharing our 25 minutes speech @ Blend divided into 3 small articles, respectively focusing on how we started working together (being a couple creatively and affectively), on how we built our team, and this last one, on how we ended up sharing our desk with a robot, able to create videos autonomously 🎥 🤖

But let’s first jump quickly in WHO is the robot, here.
So, it’s called Algo.

Yes, like in algorithm… second time, lesson learnt.

It allows us to create automated videos, or as we like to call them…

It’s basically After Effects in the cloud, working like a tireless animator 24/7

We use it to create data-driven video campaigns, like this one below, where we created one video for every medal won during the Olympics,

or creating a daily video recap of the stock markets,

or maybe a dozen of videos for every football match that give you insights about how the match is going, through data.

Here’s a really technical overview about how it works. It crunches the fresh new data every minute and, if it’s the case, it starts making a new video. It opens an AE template project, works on it super rapidly customising everything and then saves the project and renders the video.
All this in maybe 20 seconds.

But why did it joined the team?
Because why not… software has already entered the more unexpected aspects of our lives, after all:
🗣 🚕 yelling for a cab (not needed, unless you’re in 🇨🇦)
👆📱 asking a girl out (swipe right, boom!)

And honestly because a client once asked us to create a dozen of statistics videos for each of the 380 matches of the Italian Football league. Happening all during weekends, and they needed them out A.S.A.P. which for them meant in around 1 minute.
At that time we were super impressed by a project called Datagrams, from a small studio called Buck 😅, and from a bunch of other automated video experiments from a startup 😊 called Facebook.

So we said… Of course, we can totally do this!

Most of you are probably thinking how is working with Algo different than we’re all used to.

It’s very different, from forcing us to work on Microsoft servers (forgive us Sir. Johnny Ive) to finding out that’s a project that must actually work and sometimes it means having to find a bug within 20 pages of code.

But we’ll now focus on 3 examples about the design process.

1. Think of storyboards like if they were apps

When you’re dealing with a data-driven project, you have to abandon the linear storytelling flow and start building something closer to an app navigation tree.

Although in most cases scores are going to be 2 to 4, what happens to your video in a no score scenario? What happens instead if they are 12?
All this should be managed during the initial conception phase.

Plus, like in user interfaces, you should be ready to adapt your design to very different lengths of string texts, especially when Chinese and Arabic athletes decide to join the same competition, like in the Olympics.

We swear, these are real athletes’ names.

2. The design process might start in Excel

When you’re going to pull out tens or hundreds of videos in a few hours and publish them on an app or on Twitter, you want to do everything you can to maximise diversity, to let viewers perceive them as different. Using variable colour palettes is generally a great idea to achieve this.

Among our projects, we tested different options.
We started simple, by cycling the swatches of a 5-colours palette.

Then we connected different color palettes to different financial performances for Bloomberg. Medium-easy.

And then we decided to go for it: for the Olympics we coloured the videos with a custom palette in-line with the country colours of the gold medal winner. We had to create one unique palette for each of the 207 countries attending the Olympic Games.

The global result was deeply satisfying, but it was a challenging design work. We had multiple flags overlapping the coloured backgrounds, shades, texts and testing all the possibilities was almost impossible. So, using Excel was the only option. We defined the colours as variables, we used conditional formatting to visualise them within Excel, filled up all the options and then exported the palette data as a .json file, fed later to AE.

3. Keyframes are overrated — better use code

The best part about Algo projects is when the whole animation changes in function of data, not just a few texts fields and colours.
After Effects gives you complete creative freedom, allowing you to potentially connect any property of an animation.
To do so, expressions are way better than keyframes.
Most of the time we end up creating a rig, a series of expressions that connect all the elements of a scene to a single input data, so that by changing that single data, the whole scene transforms to deliver a different message visually.

We have big plans for the future 👀 of our friendly robot Algo.

Social Media are on a path to become all about video even more.
And their innovations in video distribution will also shape the kind of content we will need to produce as professionals.

Think about Snapchat, Instagram stories, disposable media there are available for just 24 hours, that focus on “today”, and also think about live video, that focuses on “now”, on this exact present moment.
There’s a clear trend going on from traditional video to these new, faster video consumption forms, and clients nowadays need to be present on those platforms to exist.
Algo is born to allow this. No matter the medium where they end up published, Algo videos allow motion design to narrate the present moment, making sure that we can help clients put their tons of data to work in their communication.

So, are robots going to steal our jobs? We don’t think so.
Creativity will remain for quite a long time a human domain.

And those who say the contrary, are probably lying 😄

But we, creative little humans, must admit there are things machines do better then us.

We think in the future, whether we as professionals like wearing nerd glasses 🤓 or not, automation and real-time video production are going to have an important role in our work, and we made Algo because we want to be part of it.

We really want motion graphics to have the chance to describe our present. And who knows, maybe also our future!