Re-Imagining Work in The Creative Age

How often do you feel creative at work? Why isn’t the whole of your day happy and playful? Technology, and particularly Artificial Intelligence, has the potential to help with all of this — but somehow things went wrong.

These questions motivate everything we do at How is artificial intelligence changing design work? What does it mean to be creative in a digital world? Why should we put more humanism into our tools and their interfaces? How can we make the workplace more playful for everyone?

(Alex Champandard, co-founder of, presenting the company’s keynote at #SLUSH17.)

This article is based on our #SLUSH17 keynote delivered in Helsinki recently, that elaborates on all these points and outlines our driving principles.

Watch the video recording of our keynote here. (11 minutes)

At, we’re in the process of re-imagining work in The Creative Age (a goal that’s as lofty as it sounds) but also redefining what it means to be in stealth! During the day, we quietly and diligently collaborate with the top graphic design agencies worldwide — bringing their creative workflows to life with Artificial Intelligence, and helping manage this transition in the least disruptive way possible.

In the evenings and during weekends, everyone on the team transforms into a casual creator in one way or another. We become musicians, poets, visual artists, comedians, game designers, performers, storytellers, writers, composers… We do it because it’s who we are and because it’s fun and messy, sometimes even beautiful — but mostly just messy!

Frankly, despite each working in the field for decades, we find today’s creative tools rather frustrating. Whether it’s off-the-shelf software or cutting-edge A.I. algorithms, it takes a lot of mechanical skill to use these tools today. The learning curves are more like cliffs, and even when you build up some expertise there’s a lot of mental strain for getting things done. It’s rewarding to be creative and innovative, but the frustration is always there…

The fact is, machines have already taken over the creative process, whether we like it or not. We create our tools, then our tools format us. The interactions and interfaces we use today were designed in the late 60s, and the software we use is based on designs from the late 80s. Our notions and entire beliefs about digital creativity have been shaped by these early decisions. Nobody wants to be formatted like this, and especially not by tools that have been around for so many decades!

Over the past year at, we went back to the drawing board and asked ourselves some fundamental questions:

  • What does it mean to be creative?
  • What defines being human, even?
  • How can we design interfaces and tools on our own terms?

Here’s what we learned… (You can probably guess what happened next!)

The first thing we discovered is that everything we do in the workplace today is wrong—at least from a creativity perspective. From our sitting postures to our office layouts, to the working hours and stress levels, the research and the studies are clear: all this makes us less creative.

If you had to redesign the workplace to promote healthy design practices, innovation and creativity, you’d come up with something that’s the opposite of what we have now!

The second thing we realized is that everyone seems to have forgotten what it means to be human. Our tools have formatted us for so many years that our entire self-identity has atrophied. Many people in the field are worried about AI because they define their skills based on the tools they use: they think of themselves as “Photoshop technicians” and somehow sideline everything else of value they bring to their studios.

At, we believe that human creators will always have essential roles in the creative process that machines would struggle to do or not be able to do at all. This mindset carries into our design process that aims to amplify human creativity. (It’s good news for all human readers here!)

Assuming there will be humans around in the future, we should probably ask ourselves the question: how do we re-design our interfaces and software based on humanist principles? For us, this has meant starting with the three universal factors that motivate and engage every person on the planet, based on Self-Determination Theory.

We’ve taken each of these three motivators (autonomy, mastery, relatedness) and turned them into corresponding principles of design:

  1. Empowerment (for autonomy)
  2. Expressiveness (for mastery)
  3. Collaboration (for relatedness)

These inspire what we do at, but they’re also principles we believe are appropriate in all other software too.


Part 1. Let’s look at the situation from a perspective of autonomy first. This is arguably the topic most passionately discussed at our team retreats!

We believe that most software built around Deep Learning is on the dis-empowering side of the spectrum. This automobile production line is a great visualization of the process: data comes in and data comes out. Everything is built around the machine, so labor is also structured entirely how the machine works: there are human-shaped holes at regular intervals. There are people to clean the data, annotate the inputs, label the outputs.

These aren’t the most empowering roles, autonomy is low because the system dictates what you do, and so it’s not necessarily very creative either. Yet this industrial age mindset is the most common today when people design their software around “Artificial Intelligence”. What would it take to fix this?

Another alternative metaphor for software is inspired by the Information Age. There are large quantities of data being gathered from sensors, cameras, trackers, monitors and it’s all presented to the users. At the same time, there are many controls available as a collection of input fields and sliders, pages of buttons and tick-boxes, as well as other actuators.

It not only requires a lot of expertise to learn to use the systems, but the bulk of information and overwhelming choice creates cognitive overload. This also has a tendency to reduce creativity too. How can we reverse this effect?

We think it’s time to flip this relationship around and make it more playful. Imagine a child playing with a toy, full of joy to the point of forgetting the world around, so deeply engulfed in the creative flow. Creators should be the ones to have the power, putting them in charge of the system as if they were its original creator. That’s how we think about the software we’re building, to make it even more empowering.

At, the metaphor we like to use is AI as an instrument. You can pick it up, perform, experiment, be creative, produce something, then put it down again when you’re done. It’s yours and you’re in control at every stage of the process!

Who says the approach you use today for your specific problem is the best choice for you tomorrow? Should anyone else make these decisions for you? Only you can determine what’s most appropriate for each situation, especially if you’re aiming to be creative!

We believe the best way to do this with software is using a modular approach where you can assemble your own tools in a do-it-yourself (DIY) fashion. This takes us back to more constructivist languages and frameworks like Smalltalk and Squeak that make it easy to modify running systems, remix ideas, and remake tools on-the-fly as you create.

When creators are in charge and can build up their own tools, it becomes much easier for them to offload menial tasks they see fit to the system. We want people to be able to dance around the automation line, for example:

  • Ask the system to generate dozens of unexplored design variations when you go for a run to review when you get back.
  • Ask the generator to make multiple variations of any particular design so it can be used in many different contexts.
  • Ask the software to perform an exhaustive review of other work in the field overnight and give you a report.

These are the good kinds of automation!


Part 2. Now let’s look at creative software from the perspective of mastery, another one of the core drives from Self-Determination Theory.

Software today is rarely self-explanatory. There’s often a learning cliff that few people overcome. Both the interfaces and the features themselves are hard to understand, and you need clunky books or awkward video tutorials to understand what’s going on!

When you do get over the learning curve, there’s a chance you’ll hit the limitations of the software. The software is usually made of complex algorithms that must be built beforehand by the original developers. At best you can hire a professional to extend your tools for you, but the software is not expressive enough to extend itself.

Algorithms for Artificial Intelligence or Machine Learning are not very self-explanatory or expressive either. Have you tried looking inside a deep neural network lately? These algorithms are built to operate efficiently on their own, but not designed to work well as components in a larger hybrid system that involves real people. With recent EU regulations on the right to explanations, this is going to affect the algorithms we use, as well as their integration into our workflows.

We suggest approaching the problem differently. At we’re designing a language for creativity that lets you express what you want to do with patterns, properties, symbols… This language can apply to architecture, music composition, visual design, storytelling, or any other creative process. All interactions are then expressed within this language!

By using combinations of patterns, properties, or symbols (whichever name you prefer for these concepts), it’s easier to see what’s happening within the system. It’s no longer a black box that takes a Ph.D. and a server farm to analyze. You know at first glance that a pattern breaks up into other patterns. Each property has a clear label and is described within the system itself.

Naturally, you can assign your own labels to these properties or symbols, those you use yourself or within your team. You can create new patterns that combine other patterns together, then start using them as part of your own dialect that you use to create things!

When you build up your own toolset and its language in a do-it-yourself fashion, what you’ve built becomes a reflection of what you’ve learned. AI then becomes a tool for self-reflection too. How do I approach this problem? What actions do I use most often while creating? What concepts am I missing that I could add to my palette?


Part 3. This brings us smoothly into the topic of relatedness in creative software, the last core drive from Self-Determination theory.

The workplace today often puts people next to each other, but it isn’t necessarily the most collaborative nor the most creative environments. The overload of information doesn’t help collaboration, the fact that we’re often seated in static positions, indoors with little fresh air, in boring and regular environments, working long hours to a schedule. The research shows these settings are not conducive to creativity.

Mobile technology doesn’t necessarily help either, even when it’s outside of the workplace. Not only are applications are robbing us from our attention, they’re preventing us from having piece of mind. It’s those moments of tranquility and calm that are most conducive to creativity. We need boredom to have the opportunity to be creative! It’s from these peaceful and tranquil moments that sparks fly. When technology gets out of the way, both personal inspiration and collaboration can happen inhibited.

What would human-computer interactions look like if we redesigned them? When you’re on a walk outdoors, why not share your ideas or suggestions immediately with the system through your mobile? When you’re in a business room, why not use the whole space and project things on all surfaces? When you’re brainstorming, why not use a more tactile approach using physical objects, scribble things down on paper and have the system read your sketches?

Thanks to progress in computer hardware and software, we can use modern AI and machine learning to help redesign these interfaces. With computer vision algorithms, voice recognition, motion tracking we can have the system operate in the environments where we are the most creative!

We can also use Artificial Intelligence to shift the burden of getting over the learning cliff over to the system too. The tools we use should be able to model us too, not just requiring us to do all the work. This makes the relationship bi-directional and moving into co-creative workflows.

Taking the next step, we can also redesign our interfaces around the idea of empowering and enlightening creators, letting them ask for help about any topic they are interested in. We can create coaches that help the creators build-up their expertise in the most constructive way. We can create experts that can explain how things are being done, for example visual design.

By using AI agents that collaborate with us, we can also build a network that includes both human creators and AI agents. This makes sharing concepts, goals, processes, histories much easier thanks to the underlying creative language that’s shared. It becomes a generative web that can produce creative artifacts!


You may have been expecting it, but there’s no product screenshot in this post. We’re not trying to sell you anything! This is about something much more important: it’s about making a choice about the kinds of systems you want to build and use in the future.

We have to assume there will be machines in the future, and for the sake of those reading this that are human, we should assume there will be humans too! ;-) The question is how do we design for the interactions? It’s all about those interactions! There are three major ways we can approach this.

We can take Industrial Age mindsets and turn humans into digital cogs. Under the guidance of strict software, these are easy roles to fill and are mechanical to perform. We can easily measure outputs and it’s the kind of product that Wall Street would be the most interested in! Of course, it won’t be very fulfilling for anyone in the pipeline, and the outputs won’t necessarily be very creative either…

Alternatively, we can take an Information Age mindset and see everyone as producers and consumers of large quantities of data. All your online tools and applications, even the office environment, is instrumented to capture as much information as possible. Having all this data available and so many choices too can make it overwhelming to the point of making you less creative too. While this approach is very popular in Silicon Valley, these open plan offices (or cubicles) are not much better than the conveyor belts. It’s easy to fall into these patterns too, but we’ve chosen otherwise.

The alternative is to think in more humanist terms, but it’s not easy. In fact, it’s the most difficult thing we’ve done at It requires everyone on the team to level up, be more mindful of every aspect of their work lives—becoming more involved and taking more responsibility. In the long run, we’ll have to change the way companies operate from the ground up to be more humanist too.

The result will be new types of organizations with a supportive culture, that embrace emergence and where team members are like musicians improvising in self-organizing bands. Not only does this require mindful design of company culture, but we also need to design our AI systems as instrument that don’t get in the way — instead empowering the creative process.

As a society, our current trajectory was heading to a dystopian future where the machines we build are either in direct control of our lives (similar to George Orwell’s 1984) or alternatively designed to exploit human nature (rather like Aldous Huxley’s Brave New World). As a side effect, our creativity and playfulness both went out of the window, but if we can recapture those there’s a chance we can build an alternative…

We call this “The Creative Age” and it starts with you! It’s your choice now what kind of systems you want to interact with in the future. Are they more like industrial pipelines or instruments to play with? Will you be mindful about these and take responsibility for their use? Not everyone will choose to craft their own future like this, it will be the hardest thing we’ve done, but there’s hope that we’re in this together!

Watch the video recording of our keynote here. (11 minutes)

Does this post give you hope? Do you resonate with our vision?

If you’ve read everything this far, we’re sure our paths will cross again on our journeys into The Creative Age. See you soon! #⚘