The world is changing. Two decades ago, we were looking at encyclopedias for information. Today, we just get out the tiny, powerful computers that we carry around in our pockets and “Google it.” Tomorrow, we’ll simply be asking out loud, “hey Google, what is XYZ?” We are living in a rapidly changing world of user interface design that’s moving from graphical to voice-based as we speak, and anthropomorphic VUI is very much a part of that shift.
There are so many things to consider when designing a voice user interface (VUI), and one of them is just how many human traits we can—and want to—achieve.
Basics of VUI design
When a developer is building an app for Alexa, Siri, and other voice assistants, there is an in-depth process known as VUI design, or voice user interface design, that comes into play. This includes determining some of the following factors as the design unfolds:
- The trigger that will prompt action, whether that’s audio, motion or something else.
- The intent behind phrases, whether they are worded as a straight-forward question or not.
- The many different ways in which a single request can be formed.
- The device and hardware used in the interaction.
- The use cases (primary, secondary and tertiary) for a VUI. Is it just used to track fitness stats, for controlling aspects of a home, or for a wide range of inquiries?
Simple VUI design alone isn’t enough. Conversations are meant to “flow,” similar to how they do with a human.
What about anthropomorphic VUI design—what’s that?
Ultimately, it is the process of giving voice interfaces human qualities—with the goal of creating a relationship between the human and machine. This manifests itself in many different ways. Is the point of it to replace humans with machines so they take over the world? No. The purpose is to create an experience so good for humans, they can barely tell that they’re speaking to a computer.
To create a truly anthropomorphic VUI, designers have to think not only about determining the intent behind voice commands and queries and providing the right information or performing the right task, but also responding in a way that feels like a natural conversation. It needs to understand the smaller words that surround the crucial pieces of information, to avoid exchanges like that of this chatbot.
It’s obvious when a conversational interface can’t keep up, and it leads to frustration. This is largely the reason that chatbots have failed to gain critical mass adoption as of yet. On the other hand, a great experience using the conversational interface feels natural. It’s just as the quote by Jared Spool says:
“Good design, when it’s done well, becomes invisible. It’s only when it’s done poorly that we notice it.”
Designers must take into consideration different ways to pick direction from a sentence like “Alexa, when is the next available flight to San Francisco?” Words like “next,” “available,” “flight,” and “San Francisco” all must be registered, understood and pieced together to discern the request. If it’s asked as “Siri, I need to go to San Francisco as soon as possible,” a good VUI will be able to discern the intent and request and return similar information.
It’s not just the voice comprehension that anthropomorphizes this technology, either.
Long gone are the days of robotic-sounding voices as the default for any AI interaction. Nowadays, we have human voices. We even have realistic-looking “digital humans” on screens and as holograms, complete with facial expressions. The physical designs of some electronics have stylized faces, and some emit light patterns that put some personality behind their audio output. Anything that can be done to make a bot seem more like a being helps in the design quest.
Anthropomorphic VUI design in action
One of the anthropomorphic features of Amazon’s Alexa is the fact that for better or for worse, it’s always listening. Just like a real person. Your little brother might be playing with legos and not actively engaging in a conversation with you, but if you scream his name because you stepped on Darth Vader’s head in the other room, it will prompt a response. Similarly, if you’re late for your flight, you might scream at your Echo to give you your updated flight info, and it will respond immediately to the voice command due to its active state.
Siri, which is kind of the OG of VUIs, has a personality of its own. There are dozens of articles online covering the personality of Siri, and people will ask many weird and wonderful things just to see how Siri responds. There’s huge amounts of programming behind the (often funny, deep, or whacky) answers it gives, but many of them do give the sense that there’s a unique mind behind the chat.
Google Assistant and “Pretty Please”
Manners, of course, should mean nothing to a computer—but Google Assistant’s VUI can be set to help users mind their Ps and Qs. Pretty Please mode rewards politely-phrased requests, which builds social skills. It’s a great asset for families with kids learning their manners, or adults who need a reminder!
#TBT: Clippy, an early anthropomorphic assistant
When I think of anthropomorphic design, I can’t help but reminisce on Clippy. Clippy was Microsoft’s word assistant, and it did have anthropomorphic qualities. It took the shape of a paperclip—but a friendly, sentient one. Its approachable appearance made talking to it more normal, and Clippy the googly-eyed paperclip started a movement—one that perhaps, through some twists and turns, has led to voice user interfaces.
The upsides to anthropomorphism in VUI
To make something “anthropomorphic” just means to give it human traits, and while it’s a word often associated with the furry subculture, it’s an important aspect of VUI design. Humans are very used to interacting with each other. So, if we want to have seamless conversations and exchanges with bots, they need to be as human-like as is, well, humanly possible.
Effective communication is a huge asset for businesses, whether for customer support or sales, and it’s ideal that they don’t have to employ a huge number of real people to make it happen.
Increased anthropomorphism in a voice user interface also helps to build trust between a user and the product. Trust is the holy grail for brands, and if they can offer an interface that instills confidence in their customers, they are on the right track. People are unlikely to trust or form any kind of relationship with anything that presents entirely as a computer, but one that feels a lot like a living creature and can interact on their level? It’s a powerful tool.
What to watch for
VUIs designed with anthropomorphism in mind are great in many ways, but there are some points of friction to consider when using voice interfaces in your product or service.
The creepiness factor is a big one. Some people just don’t love interacting with AI, and they especially don’t like it when that AI pretends to be a human. There’s not too much to be done about that, as you can’t please both camps. As chatbots and VUIs become more popular, this may lessen. Creepiness is also attributed to personalized advertising which tracks your habits, and this can come into play when you combine voice interfaces or chatbots with machine learning that allows them to “get to know” a user.
Even if they don’t mind talking to robots when there’s a clear purpose to the interaction, excessive words and frills can put some people off. In a lot of cases, people are conversing with technology with a very specific goal in mind—and the anthropomorphic design shouldn’t get in the way of that. Think of a cashier that’s too chatty and slows down your grocery shopping experience. It can be frustrating.
Can we be friends?
While most people don’t really feel the need to be friends with their voice assistants, chatbots, and other voice user interfaces, it’s nice to have a friendly interaction. More importantly, it’s nice to have one that makes sense and achieves what we need it to achieve.
Botmock can help you to create chatbot conversations that provide just that—smooth and effective interactions with your customers and clients. Schedule a demo today to find out just what our powerful platform can do.