Is AI the next big thing for Design Thinking? (GPT-3 Series #1)

A group of Italian Designers is leveraging AI power to remove frictions from the innovation process. The results are astounding.

Photo by NeONBRAND on Unsplash

The article’s core idea and about 50% of its words were written and thought by a machine, starting from the keywords: “Design Thinking, AI, innovation process”. About 9 iterations were made to find their final form. The remaining part was integrated by our business writers in order to give context to the algorithm peculiar insight.

Follow us on LinkedIn, where we are publishing an articles’ series that shows the potential for a new type of interaction between AI, Businesses and Design.

To know more about our GPT-3 Experiment, click here (ITA).

For many businesses around the world, continuous innovation is existential.

This pandemic-related crisis just proves that.

But Great Innovation doesn’t happen by chance.

That’s why Design Thinking has become a mainstream synonym for innovation management. It’s because it can be one of the most powerful and cost-effective tools to facilitate innovation processes. It is used by organizations of any kind all around the world, from small businesses to giant companies to political entities.

Does it really work?
Yes. And no.

In the last 20 years, Design Thinking has repeatedly shown that it can produce great results in so many fields.

But so many results of these innovation routines are still underwhelming.

For many of companies, what it is labelled as a Design Thinking approach is the same old waterfall, top-down, approach to R&D. With a sprinkle of empathy, a zest of what-is-perceived-as tech, and some critical thinking moves.

But it simply doesn’t work like that.

Why are we leaving money on the table?

Why this approach’s potential is so under underutilized?

Design Thinking (as the name states) it is not only a tool. It is a way of thinking. It is a sort of operating system for your organization — and your brain.

The main problem is that many organizations are static, and there is no real space for continuous innovation, the kind of which the current market condition demands.

The second problem is that Design Thinking requires… thinking.

And thinking is hard.

Let me explain: the process in itself it’s anything but hard. Everyone can participate in a workshop or a sprint. Everyone can contribute to the result.

That’s the beauty of it.

You don’t require special skills. You don’t have to learn anything new other than what you already know.

The entire thinking process is carefully designed to make the generation, discussion and improvement of ideas the fastest, the easiest and the most cost-effective possible.

But that’s not the end of the story.

And you know that.

Thinking together is even harder.

When you gather different people around a table and you try to align them towards a goal, there is the potential for a lot of friction.

All kind of psychological, social and cultural problems arise. Of course, as Designers, we’re well equipped with skills and tools to smooth out the interactions.

But the root of the problem can still be there.

Idea generation is hard. It drains so much energy. And it depends on the personal disposition and capabilities. For someone, just explaining an idea in public could feel like a 90-minutes football match.

And it requires time. Multiplied for every participant.

And time costs money.

Of course, it’s an investment. A single workshop can jumpstart the next big idea that will make money for years to come.

But the innovation process is not a linear one. Sometimes years are needed to transform ideas and prototypes into business cases and realities.

The process often succeed. But it can fail. Courage is needed, because the outcome it’s not always predictable.

Often the solution appears serendipitously.

And, as much as Designers love the idea of controlled randomness that saturates every workshop, business owners usually think otherwise.

And often it is really hard to sell that idea of Creative Confidence as the “most cost-effective way to innovate”.

The idea generation is the most fragile process.

Because it relies on people. Sometimes it requires a lot of people to be fruitful.

And a lot of people equals a lot of noise.

A lot of things can go wrong. A lot of people needs to be in the same room at the same time.

A lot of things have to happen simultaneously, and the current pandemic proved that it’s not always possible.

Furthermore, people are dynamic entities.

Even genius people (the kind of people the Design Thinking process tries to elevate us to) can not feel well.
They can have other problems occupying their mental bandwidth.
They can sincerely not have a single good idea for an entire workshop.
They can not be having the soft skills required to collaborate.

And usually, that’s fine: the process just works anyway.

So… how the process become even better?

In order to become better, it has to stop relying exclusively on human-powered ideas. And words.

Until some years ago, that was impossible. Even with the most advanced tools available, the only contribute machines would bring to the table was as information containers (and not even the best kind: we prefer post-its) or some useless random word generation.

The good news is that we now live in a world where AI deserves to sit at the table.

Thanks to models like GPT-3, we can now outsource part of the thinking process to the AI. And the results are, in most cases, indistinguishable (if not better) from those human-generated.

Think this: if the quality is the same, and the quantity can be infinitely (almost) scalable, the process as a whole became faster.
It is like sitting at a table with 100s humans willing to cooperate, without all the noise.

How is this even possible?

Without entering the details, it is just possible to say that AI is trained on human materials, human ideas, and human-generated data, and so it can output the same kind of material, but connected in an always-new way.

After experimenting for some months, we at Lato have now the tangible proof that bridging synthetic intelligence and human creativity in order to create better experiences for humans it is possible.

It’s not easy, yet, but it is possible.

Even for smaller teams, even without dozens of computer scientist.

Imagine the AI as one of the designer sat at the table. It’s not the most important one. It’s just A designer.

In the same amount of time a human can generate 8 crazy ideas, the AI can run hundreds of humans-like user experience simulations in parallel. Imagine gathering in matters of seconds a big quantity of specific insights about that particular argument, from multiple points of view.

Quantity-like, it is simply impossible for humans to produce the same amount of synthesis without hours of research and alignment.

It sounds fantastic. And it really is.

But there are concerns.

Innovation is never neutral. Things that always have a predictable and an unpredictable impact.

That’s why we have to be careful. We’re designing a set of internal policies and solutions to address the most concerning use cases. Here just two of them, but there are many more that we, as a Benefit Corp, are carefully exploring:

1) The AI doesn’t think. At all. It just produces probabilistic output based on a specific input.

The same input, given two times, will generate different outputs.

For example: given the prompt for a product idea: “an ice cream that doesn’t melt on your hands” the outputs (selected between many) were :

“a cone with retractable folding lips that catch the melting liquid”

“socks for hands that can be washed after every use”

“an ice cream that can be inserted directly into your mouth”

Those ideas look silly. But that’s the process generation ideas. Human ideas born during a generative session are not more serious or concrete than that.

By the way: have you noticed that the third idea is much similar to a recent innovation of the ice cream industry, the mini icecream?

What if in a workshop the AI spit out a line like that, and humans start building on that idea, and a prototype is born?

Idea generation is never the finish line. It’s always the starting line. Ideas need to be refined. The process itself it’s an iterative one. But you have to start somewhere.

What’s the problem then?

Humans can declinate any ideas according to a purpose.

For good and for bad. And that’s the problem.

2) Every output of the algorithm is potentially toxic for someone or something. The AI doesn’t know anything about good or bad. It is not aware of the implications of things said and done.

Models like GPT-3 studied from content that is publicly available on the internet. It can easily say contain pieces of information that are sexist, racially offensive, or plainly wrong. And are certainly biased since they’re the byproduct of a certain kind of culture.

Humans are ALWAYS required to be guardian, fact-checkers and impact evaluator.

Is it the end for human designers?

No, quite the opposite.

It’s now easy to imagine a future in which humans will be free from time-consuming tasks like generating the right words at the right time to explain a concept.

This time has never been so near.

Of course, there will be an enormous revolution that will affect all fields of creativity.

Humans are debating if will it be a liberation or an enormous loss.

But the same thing was said when calculators were invented.

Did the invention of calculators mean game over for mathematicians, accountants or engineers?

Not for all of them, and — in the long term — not at all. It freed the mental capabilities that we now use to do bigger things, faster.

Humans are still required to build a bridge, but they now serve a different role. They don’t do pen & paper calculations about forces and materials. They now evaluate the results calculations made by a computer; they modify the parameters of those calculations, and they judge what’s better to reach the goal.

To put it simply, we don’t do the “how much…?” anymore. We just do the “Is it good according to our goal…?”

In a near future, we won’t be asking anymore “how can I explain this…?”, we will ask “is it a good idea? Who will benefit from this? Who (or what) will be damaged by it? How can we maximize the first and minimize the second?”.

Everyday we use similar shortcuts that free us from time-consuming tasks.

In order to explore a city, we don’t need to study the city’s street in advance. There’s Google Maps for that.

In order to commute every day with our car, we don’t need anymore to have our hands on the wheel. There’s a Tesla for that.

In a matter of months, not years, we will outsource (responsibly, we hope) one of the most time and glucose-consuming task since our species can perform: align everyone around the table around an idea. Soon, there will be a machine that generates better words than any human can do.

Lato is an Italian Benefit Corp obsessed with delivering great Customer Experiences that serve profits, people and the planet. We help businesses build better experiences by helping them understand more about their customers. We leverage that knowledge for a greater good with the most advanced technologies available.

Learn more about this series by clicking this.

Visit our website by clicking this.

--

--

--

Lato is an Italian Benefit Corp obsessed with delivering great Customer Experiences that serve profits, people and the planet. We help businesses build better experiences by helping them understand more about their customers. We leverage that knowledge for a greater good.

Recommended from Medium

Beware of chasing the dangling carrot in front of your nose

3D Building Model in SelfCAD

Brief Note on SketchUp LayOut

Assignment 2 — HPW

How to Search For a Permit, Report a Code Violation and More

Empowering Product Teams by Scaling up Design Sprints

Accessible content is (a) right

A laptop screen with the text designers should always keep their users in mind

5 design essentials that makes a design aesthetic

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Gianluca Amato // Lato Società Benefit

Gianluca Amato // Lato Società Benefit

Writer and Semantic Specialist that designs people’s interactions between Words and Experiences. Lato Benefit Corp Co-Founder.

More from Medium

Product Innovation Through Collaborative Intelligence of Human and AI

ARE CHATBOTS USEFUL FOR THE HR DEPARTMENT?

The People Behind The Bots — Shalini Johar

AI in Medical Documentation: Solving Physician Burnout and Improving Patient Outcomes