Design Transformation [Act 01]

This is a curated transcript from a presentation at the Machine Learning Labs talk series at Microsoft in May 2020. It is being shared as a contribution to ongoing efforts to understand and define the future of design and its transformation narrative.

(Addressing the audience in the ML Labs session)

Hello everyone,

I hope you enjoyed the topics presented so far, in fact, Margaret Maynard-Reid and Sergey Nefedyev and I have been soft-collaborating on some of these topics because they are relevant to what I do at the CSEO Studio.

CSEO (Core Services Engineering and Operations) is driving Microsoft’s digital transformation. We set up a studio about three years ago, and since a year ago I have been collaborating with a team we call Studio Transformation. We are trying to scale design beyond the walls of our studio. Scaling design is design augmentation. We conceived and run a program and campaign called CREATE TOGETHER, and we have designed and curated a portfolio of activities within the program with the aim to cultivate curiosity, capability, and community, and empower everyone to design within engineering. It is a UX transformation initiative, and we're trying to understand how design and UX are transforming. 

Today's topic though goes beyond traditional UX. I want to focus on design transformation in general and why I think it's important. UX design is just a subset of design with the latter encompassing a very broad range of skills. And as I like to try and probe the future about how we're going to create, in design, there is this transformative aspect that is happening all around us and it affects design as well. Thus I think it is imperative to think of design transformation all up, rather than UX design alone.

If you reflect on what we used to do, we humans, when we create and design it used to be something like this. 

It used to be very direct. For example, I have an idea of an object to design, or something I want to create, and I have some materials and then directly with my hands, some tools and some process I go from the idea to the creation of something, which is an object coming into existence through a nearly direct path of creation; from idea to existence. It is linear and natural.

Perhaps what I am creating is something practical and useful, or it is being made for aesthetic reasons. 

With this simple example I want to go back years, maybe even centuries ago, where creation used to be very natural, direct, and human. And if you observe what has been happening during the last years, we see more and more interfaces coming in between our ideas and the final outcome. Yet we have a long history behind us where humanity has been creating and inventing in more natural ways than today.

And there are other similarities that I was trying to understand, for example how in other areas they've gone through such disruption or transformation. I was thinking an analogy is how you pilot a modern airplane, let's say Boeing and Airbus the two dominant airplane manufacturers around the world. 

Similarities in a technological context? Boeing and Airbus

So, what do you see here? It is a Boeing cockpit. I believe it is from a 777 and you can clearly see that the pilots have in front of them the control wheel. And Boeing has this principle to trust more the human pilots, they can supersede the Fly-By-Wire (FBW) automatic commands by exerting enough force on the controls. It's part of their philosophy to have the system designed as such. Of course, the cockpit is packed with the latest technology and advanced avionics, yet there is an emphasis on the physical interface.

Then, Airbus a few decades ago started revolutionizing that model by emphasizing the Fly-By-Wire (FBW) model which started eliminating some of the direct controls and replacing them with digital inputs to commandeer the aircraft, and minimizing the main physical controls which became the small joy-sticks to the left and to the right on the picture. 

In Airbus' model, the trust is put more on the FBW system: if the system is fully functional ("Normal law") it will always have priority over the pilots to protect the aircraft and keep it in the envelope, but they still have room for "non-standard" maneuvers. Now, imagine if you had to design something in such an indirect way.

And of course, in between these 2 paradigms, there is still a lot of input required, and there are many other controls in the overall context of piloting such a complex machine as a modern aircraft. Indeed, it is the complexity and sophistication of the technologies in modern aviation that require a layer of simplification for humans to be able to interface with such systems, and the result is that through limited controls we are able to commandeer very complex systems.

Meta-thought: Imagine design becoming more inter-mediated through technologies that are removing the ‘direct design controls’ we used to have, and creating additional transmission lines in between thought/idea and design output. This fundamentally starts to require different processes as well as different skills in how we design from concepts to execution. 

I bring this analogy between the cockpit design evolution case and designing to help us see the direction we are heading in design. UX design and design in general - by using data, artificial intelligence, machine learning and whatever other technologies will exist in the future - is becoming more and more intermediated. We are being distanced from the conceptual creative instinct in our minds, the moment of ideation, to the final design outcome, through the emergence of new paradigms in how to design which are more esoteric and difficult to directly control. And in some instances also impossible to comprehend because of the computational nature of some of these realities.

I've been trying to deconstruct that space and create 3 models to help me understand what is happening. They are hypothetical at this point, and although it is an ambiguous and risky move to make certain statements, I think we still need to bring them forward as a point of discussion. And in fact, these ideas require discussion, not only among designers, but we need to broaden the discussion among larger audiences that in the future will represent the designers of tomorrow, for the design of the future. 

I see three basic stages in design transformation.

The decoupled model is essentially the model of yesterday, and perhaps a bit of today, and I think we're all familiar with it. 

In this model, design lives in a traditional state, where essentially there is a complete output from designers that is being integrated into a computational creation, which is a piece of software. But of course, the same model can apply to other domains in product design. In this case, the design is primarily generated from a designer or a team of them collaborating, and then that output is being integrated into a more complex entity, and there is a linear product creation process that can repeat. 

In the decoupled phase, we understand the relations between design intent and outcomes and directly execute actions generating a design.

For example, I want to design something in this way, perhaps I have scenarios, I go ahead and draw something and that is the design. I directly design it. I mentally see how it works and then I generate the design. And that's pretty much the world of yesterday and a bit of today.

The next phase is the transitional phase which I also call the transitional design. Maybe this is a bit the Airbus case if we revisit the cockpit metaphor, where we start seeing the introduction and emergence of other design skills represented by non-human capabilities such as AI/ML and in general new technologies that can generate design and live within this computational creation space.

So, suddenly there are other types of designers 'out there'. They're not human designers, and they're capable of designing 'small things' now. They are leading that immaterial space in design as our practice is transitioning to a new phase. Probably most of the design activity is still outside in the traditional dimension, yet there is this emerging and growing new reality in design that manifests itself within the computational creation, which is outside our traditional envelopes of creation.

In the transitional phase, we mostly understand, and sometimes we directly execute the actions that generate the design. 

Many, if not most, designers are still working in the traditional, classic way. Meanwhile, design transformation is already undergoing and is in the initial stages of a longer wave and journey. We start experimenting with these artificial capabilities of design, and probably we don't understand exactly how they work. And we don't know how to put all things together into a new clearly defined process. It's a very abstract space.

And the next phase, which is the third one, is the one that I think will be more disruptive, and again it is a hypothesis. 

The intermediated design is where things are fused. If you notice (in the diagram) there is a significant change with the design discipline. The role of the designer has changed. I've changed those circles and the designer titles, but the titles are just examples for us to think about what those future roles will be. Imagine a situation where most of the creation lives within that computational space that has a lot more capabilities in design, and it is self-generating, perhaps even self-driven, and capable to find more suitable solutions than the designers. And if we let it – make also design decisions on behalf of us the designers, the creators. 

In the fused phase, designers become orchestrators and trainers of the system that generates design.

We as designers will be of course collaborating with that computational creation, but then we won't be just designers. We will either be design orchestrators, design philosophers, or embrace new roles that we haven’t yet imagined or invented. Because we went from a legacy situation where we had an idea and with our hands and tools were creating something, to a situation where the creation of something is being driven by a new amalgamation of the means of creation (call it data, new knowledge, technological abilities that constitute other forms of designers, new tools, etc., they are all next-generation interfaces in the process of creation). And we, as designers, need to move to the next level to find new ways to collaborate within the fused model.

These 3 models are obviously not complete, but I use them as a reference to bring this conversation forward and to understand both myself and also in my conversation with others about how this evolution potentially may take place. They serve like early blueprints to raise more questions since we do not have all the answers at this point.

This was an introduction to how I see the transition of design to its next phase.

In Design Transformation [Act 02] I will introduce my thoughts about a model to de-construct and roadmap design transformation across four phases that define the transition from the past to the future.

Thank you,

Yannis Paniaras

Microsoft Design & CSEO Studio