The Rise of Artificial Intelligence: How AI Will Affect UX Design
Artificial intelligence is no longer confined to the domain of developers and data scientists, says Josh Clark, founder of Big Medium, author of several design books, and the inventor of the Couch 2 5k program.
With an unprecedented amount of data being collected and algorithms driving many of our interactions, Clark says the challenge of what to with the data, how to present it, and how to use it to shape user behavior is now inherently a design question for today’s user experience designers.
“Typically digital designers have been creating interfaces for flows and for content for which we have complete control, and now we are gradually and slightly uncomfortably feeding some of the control to algorithms and digital models that we don’t quite understand—and that don’t quite understand us,” he said.
This Isn’t Skynet
AI, especially in terms of voice interaction, has skyrocketed in capability, but it has also brought with it a number of challenges for designers.
“I think there’s often an assumption or a fear about artificial intelligence and we quickly go into the Terminator mode that Skynet is becoming sentient,” Clark said. “That’s a long ways off.”
Clark says more immediate concerns for designers include:
- Figuring out how to address user frustration when voice-enabled products like Siri, Alexa and Cortana don’t fully comprehend what the user is asking
- Anticipating and designing for experiences that aren’t yet capable of being fully seamless
- Dealing with the pressure of having to execute flawless, cutting-edge AI experiences at both a client and consumer level
“I think the really critical question we’re dealing with now in this early stage of artificial intelligence is how do we design the interfaces in ways that set appropriate expectations and channel user behaviors in ways that match the capabilities of the system?” Clark said. “When expectations are wrong or we ask the machines to do something they’re just not capable of, that’s frustrating, and sometimes damaging.”
The Biases of AI Must Be Challenged by User Research
The damages he is referring to are the unanticipated biases of machine learning. He points to examples like algorithms that show racial bias when they are unable to recognize or detect faces that are not Caucasian.
Clark says “UX research at an unprecedented scale” is the solution. It is the only way to begin filling in the black holes caused by all the unknowns. Just as machines can fill in gaps for humans, humans must fill in the gaps that machines aren’t able to understand.
“If we want things to work for everybody and not some narrow average then we need to open up to a volume and diversity of test cases in order to make sure these things work well,” he said. “That’s true in general, we’ve just never before had the capacity to embrace anything beyond a relatively narrow dataset. Now we can through data processing and the wealth of data that is out there literally try ideas, concepts, products and certainly data models on millions of people.”
How Human Do Users Want These Products To Be?
Another line designers must dance is down in the uncanny valley, “a concept where the more humanlike it is, the weirder and more off-putting it is until it’s exactly perfect.”
AI is bringing up a lot of questions that we’re not sure how to answer yet. How much personality are systems like Siri or Alexa supposed to have—and how much do we want them to have? Are they merely utility or is AI allowing our relationship with technology to deepen at an unfamiliar level? Do we want to connect with our computers like the lead character does in the Spike Jonze movie Her?
“I think there’s a lot of unknowns in this right now. There is an instinct that the more that we can communicate with machines on a human level rather than on a machine level, the easier and more effortless and more convenient it will be,” he said. Over time, user interfaces have evolved to take on more human qualities, including the sense of touch and now the ability for to speak verbally to one another.
“There seems to be this constant evolution towards dealing with machines on a human and even physical level as much as possible because that’s the way that we think and communicate, but it starts to become distracting when the machines don’t understand, or understand incompletely.”
Will AI Take Jobs Away From UX Designers?
Clark recently blogged about an article that ran in the Atlantic about GoogleRNN, a system that is teaching machines how to draw. These experiments, Clark says, explore the relationship between creativity and art with intelligence. He mentions to RNN’s drawing of a cat. Although the machine cannot understand what a cat is, it can still sketch it and comprehend that it is a symbol of a cat.
“At the moment, these experiments are mostly just mimicking what we do. It’s watching how we sketch and returning it back to us. What I think is interesting is we’re starting to teach these models familiar symbols,” he said.
Based on this, you could visualize these systems being trained to interpret effective user interfaces for different needs. Clark gives Pinterest-style browsing experiences, ecommerce experiences and media reading experiences as examples, and says that these systems could begin to identify common patterns for solutions that yield different goals.
“If it understands what those layouts are and it understands what symbols within those mean, you could begin to imagine systems in the relatively near future starting to essentially sketch out interfaces or wireframes for us, creating at least a first draft based on information that we feed them,” he said.
It is not impossible then that artificial intelligence will soon be capable of low-level production work that is comparable to that of a junior level designer, he said. This introduces the question of does this take away jobs or, “does that then free us in useful ways to think about our work in more creative and expansive and perhaps more strategic ways than just tactical strategic aspects?”
Rest assured, Clark doesn’t feel like designers have too much to worry about because at the end of the day machines do not think like us. He references the work of designer Matt Jones who has written on how AI should be viewed more as a companion species, and that of Kevin Kelly, the previous editor of Wired magazine, and his thinking on how AI’s true power lies in its alien intelligence that is separate from ours.
“It’s really not something that would really ever replace the way we work or think, but becomes an interesting companion to how we work,” Clark said. “I think that’s especially true in the creative industries where it can be this useful ride-along sidekick to make our work hopefully better and more insightful and let us focus on the stuff that we can uniquely do as humans.”
AI is Changing The Game
AI is asking more of designers and Clark encourages designers to begin exploring AI through the various APIs that are available for free online in order to gain a better understanding of both the potential opportunities and obstacles AI presents.
“I think part of our new role is designing not just the interface, but anticipating the weirdness and ambiguity that can come back from these systems,” Clark said.
“I’m incredibly enthusiastic about what were going to be able to accomplish and do with machine learning and artificial intelligence, but I also think we need to treat it with scepticism and critique as we begin to integrate it into some of the most fundamental parts of our society and our culture.”