Connecting artificial intelligence with digital product design
TLDR: If you’re a designer, make one summer promise for yourself – familiarise yourself with AI over the summer. Read an article, read a book or go through some case studies. You will thank yourself later.
A new article about AI? Again?
Yes. AI can claim it’s title as the Buzzword of the Decade already. There simply isn’t an industry left untouched by this intriguing, mysterious and even scary megatrend – at least as far as prophecies, general hype and seminars go.
Now, it’s design’s turn to listen.
We’ve heard the talks and predictions. It’s time to see what AI really entails for our line of work? What has and can be done? I had a very genuine human chat on artificial intelligence and design with Vilma Sirainen, a Senior Product Designer at Zalando.
Human: What do we want!? Computer: Natural language processing! Human: When do we want it!? Computer: When do we want what?
Yes, very funny. The computer is silly. But if artificial intelligence is anything like what it’s cracked up to be, it’s not simply a laughing matter.
In the gag above, the point is the folly of machine learning. There’s another term. Indeed, machine learning and AI seem to be interchangeable in today’s LinkedIn posts and water cooler discussions around town. For product design, the key interest is in machines and algorithms using existing human data in creating something new, something more personal or something better. Something that would require unreasonable amounts of human resources or maybe even something our petty human brains simply can’t come up with. Or then just streamlining some parts of the designer’s workflow and reducing recurring manual tasks.
Does the general buzz, enthusiasm and speculation mean AI is actually meaningless? Far from it. And designers should pay attention because AI has already been harnessed in our field. We’re also making use of foody algorithms and behavior-based predictions in our product, but more about them in the near future and let’s focus on fashion here.
Fashioning the definition of AI
Most of you must be familiar with AI’s points on a conceptual level. Let’s not waste too much time on that and cmd+c stuff from our friends at Merriam-Webster:
Artificial Intelligence (Definition)
1. a branch of computer science dealing with the simulation of intelligent behavior in computers.
2. the capability of a machine to imitate intelligent human behavior.
(Also, James Cameron says it will eventually try to kill us all.)
Fashion recommendations by humans – or robots
Vilma is currently working with some rather interesting AI stuff at Zalando, which is known as Europe’s largest online fashion platform. Remember how I hinted that designers should buckle up? Because at Zalando, AI is a part of their design teams’ daily work already.
The company has even conducted their own variation of a Turing test by showing fashion experts and other fashion-related people outfits to rate and comment on.
The twist? The raters didn’t know which outfits were created by humans and which by algorithms.
Direction of AI development
During our lunch moment, Vilma sketched this neat model (above) on a napkin, depicting the movement of actions in the field of AI; automating custom-made offerings for individuals.
1. Manual + for individual: Hand-picked for you
“A person will help you find exactly what you want.”
2. Manual + for everyone: Hand-picked for all
“Here’s what’s popular/in season.”
3. Automated + for everyone: Auto-related to this
“If you like this, here are more like it.”
4. Automated + for individual: Auto-relevant for you
“All of these are tailored to your specific likes and in-the-moment goals.”
Usually, number 4 is the sweet spot for growth companies. Offering relevant, personal, up-to-date and maybe a bit predictive content or recommendations based on many variables by using user data,
demographics, psychographics and other ways to recognise the needs of the customer.
Some might say that this kind of approach will inevitably lead to boring, math-based end results because human factors and phenomena are forgotten. True. Unfortunately often, that’s the case. However, making custom solutions and being able to scale them are not enemies with each other. It’s about how we build and design our algorithms so that we can consider also human values, opinions, attitudes, interests and activities. I’d say clicking nicely targeted CTA’s (or ads at all) are not at the core of the solution.
Instead, by using already existing models and well-made implementations as a source of inspiration, we have some very promising possibilities in the field. To get some concrete knowledge and actionable tips, please jump into the next section.
Learn before they do
Are you a designer and interested in AI? Vilma and I gathered a neat list of interesting reads on the topic. The machines are already learning and so should you.
Start from these
- Case studies: Algorithm-driven design – how artificial intelligence is changing design
Great collection of case studies on how to make use of AI in developing digital products.
- Free online course: The elements of AI
Free online course on AI by Helsinki University and a software company Reaktor.
- Microsite: AI Playbook
This microsite is intended to help newcomers (both non-technical and technical) begin exploring what’s possible with AI.
- Book: Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots
A good read about finding the common ground between humans and machines.
- Podcast: Data Skeptic
A weekly podcast featuring short mini-episodes explaining high-level concepts in data science.
Art and fashion-related AI inspiration
- Article: The Relationship Between Art and AI
- Article: Will AI kill creativity? (the answer is no and fashion designers give reasons why)
- Article: Artificial Intelligence Empowers Designers in IBM, Tommy Hilfiger and FIT Collaboration
- Article collection: AI and the future of UX
- Article: Machine learning guide for average humans
- Book: Algorithms to Live By: The Computer Science of Human Decisions
How computer algorithms can be applied to our everyday lives.
- Case study: How Netflix personalizes artwork
Have you ever thought who the heck is so patient that they can photoshop every movie thumbnail for every size and format and also for different markets with local title translations? I won’t give an answer, but this article is a bit more helpful.
That’s it! We’ve gone through the definition of AI, a practical example of its benefits, a possible direction of AI development and listed some concrete learning resources. Hopefully you found this article helpful. If you’re looking for immediate action points, check out one special suggestion below.
Psst… there’s an event about AI and design coming up next week in Helsinki!
Have you wondered how AI will affect the work of designers? Will AI make the design process easier? Can we apply design thinking to AI development? These topics are discussed in the first ever Dash Design Conference alongside a Saturday brunch.
The event will be organized on Saturday 16.6. at Maria 01 in Helsinki downtown. In addition to keynotes and a workshop they will be serving a small brunch, which is quite nice.
Check out the event on Facebook and be there!
From trends to methods and inspiration – 9 types of design resources we love at Wolt
We started the search for design interns at Wolt a couple months back, and wouldn’t have believed what followed. This article is our way of humbly thanking the awesome people who applied to work at Wolt.
Shifting from userless design to people-focused collaboration
I am a designer at Wolt – but let me stop you right there. While I work at the hip and cool food delivery app startup, my responsibilities lie somewhat in the dark.
Shipping without bullshitting – design intern Kerkko has something to say
It was supposed to be a regular chat with a new staff member. It ended up being about onions, crappy design and the definition of good ol’ BS. Read on to learn how to minimize bull and why design alone isn’t magic.