We stand with Ukraine and our team members from Ukraine.

The Community Of Over 640,000

Home ›› Conversational Design ›› Conversational AI Experiences Don’t Have to Suck

Conversational AI Experiences Don’t Have to Suck

by Josh Tyson
11 min read
Share this post on
Tweet
Share
Post
Share
Email
Print

Save

A Q&A with celebrated tech leader and design pioneer, Robb Wilson

In this candid conversation, celebrated tech leader and experience design pioneer Robb Wilson talks about the lifelong journey in technology that informed his new book from Wiley, Age of Invisible Machines: A Practical Guide to Growing a Hyperautomated Ecosystem of Intelligent Digital Workers. Equally at home reconfiguring code or designing conversational flows as he is building a house or carving his own surfboard, Wilson’s breadth of experience has proven invaluable in a sprawling and complex space.  As the founder of OneReach.ai, his work has come to define the conversational AI marketplace, but there’s much more to discuss. Before founding OneReach.ai, Wilson was already  the owner of UX Magazine, and here he explains how this publication fits into a much larger vision of technology not leaving people behind.

When I came on as managing editor of UXM back in 2012, Robb was busy running multiple startups and divided his time between Denver and Kyiv. I rarely saw Robb, but his work in both technology and design are legendary and he was always there, so to speak. A few years ago I jumped at the opportunity to spelunk the illusive Robb mind for a white paper that grew and grew until it became Age of Invisible Machines. Co-authoring a book with Robb came with the good fortune of many long and illuminating conversations that tended to change my perceptions of the world—the following exchange was no exception.

Your career began in the film industry. How did that influence your later work?

Well, In the early ‘90s, I worked in sound design at Warner Bros., where many of my colleagues had been in the film industry for nearly half a century. At the time, the film world was transitioning from analog to digital. I realized that the wisdom these sound editors brought to their craft was in danger of being lost if the industry started hiring people who were simply better at using computers. There’s a very distinct rhythm to editing film, and a lot of that comes from the physical act of cutting and splicing celluloid, so I designed a digital editing interface that mimicked the controls of an old school sound editing system. The older, experienced editors were able to acclimate to the digital editing process and younger editors learned the rhythms of working film by hand. The experience showed me how important it is to bridge the gap between humans and machines.

From film you moved into experience design, where you enjoyed considerable success. How did that shape your perspective on conversational AI?

As you know, I don’t like waving my accomplishments around, but it’s safe to say that almost everyone on the planet has touched a piece of technology I had a hand in developing. I was tapped to design one of the first iPad apps and, over the years, I’ve worked with many of the largest technology companies in the world. I’ve also collected over 100 awards, which I only bring up because roughly half of them are in technology and the other half are in design. This balance is important, because while conversational AI involves the orchestration of a whole host of complex technologies—like NLU/NLP, code-free design, RPA, and machine learning—its success is entirely dependent on adoption, which can only be achieved by focusing on experience design.

PQ: “I didn’t want to see AI leave people behind, so I decided to do something about it.”

I’ve noticed that  balance between technology and design is central to a lot of your work.

Absolutely. I’ve been exploring conversational AI for over two decades, and what first drew me in was the fact that, of all the experiences people had with technology, conversational ones were uniformly the worst. Once conversational experiences stop sucking, however, adoption will skyrocket. Conversation is an interface that requires zero training, which gives anyone, anywhere the ability to leverage powerful problem-solving machines. Basically, I didn’t want to see technology—specifically AI— leave people behind, so I decided to do something about it.

PQ: “OneReach.ai scored highest overall in the Inaugural 2022 Gartner Critical Capabilities for Enterprise Conversational AI Platforms report.”

It’s been a long and somewhat fraught journey, but there have been some huge milestones lately. The biggest is that OneReach.ai scored highest overall in the Inaugural 2022 Gartner Critical Capabilities for Enterprise Conversational AI Platforms report. For me and my team, it validated our philosophy that the only way to really succeed with conversational AI is to have the ability to orchestrate all the associated technologies using code-free tools on an open platform that’s flexible enough to incorporate the best products that the marketplace has to offer.

Are there common mistakes people make when trying to design conversational experiences?

I think one of the biggest misconceptions people have about conversational AI is that the goal should be to mimic human interactions and behaviors. The real power in working with conversational AI is to orchestrate the associated technologies in ways that create better-than-human experiences. Mimicking human behavior can be powerful when it comes to entertainment applications, like gaming, but in productivity settings it’s much more impressive to have high functioning bots—in the book we call them intelligent digital workers, or IDWs—giving people direct and easy access to seamless integrations of technology. The conversational interface exists to obscure all of the insanely complicated sequencing that’s going on behind the scenes. That interface doesn’t need to be clever or connect with users on an emotional level, it needs to help them accomplish more by automating menial tasks in innovative ways.

Gartner has said that something like 90% of all conversational AI applications out in the world are going to fail, if they haven’t already. Why is this the case?

It definitely ties into what we were just talking about. There are many fundamental misunderstandings surrounding conversational AI, a situation that somewhat mirrors the way companies approached UX a decade ago. Everyone knew they “needed UX,” but very few organizations really understood what that meant. Many UX efforts stalled out at the point of bigger buttons and hamburger menus either because companies failed to realize that creating a dynamic and rewarding user experience requires a coordinated effort, or because that realization itself seemed too costly to contend with. Companies that take the bigger buttons approach, in this case throwing isolated bots at individual workflows will quickly hit a ceiling. Like experience design, conversational AI and hyperautomation will require a fundamental restructuring of most organizations around building an ecosystem where bots, functionalities, data, people, and things, can thrive together. At this point, the companies that are hitting a stride with conversational AI are the ones that were built around this kind of technology. The insurance company Lemonade is a good example. They’re relative newcomers to the space, but have already severely disrupted existing business models by providing users easy access to various policies through tech-enabled experiences. People enjoy interacting with their intelligent bot, Maya, because it runs background tasks efficiently while providing users with a clear and rewarding conversational experience. They also have an employee-facing bot that’s part of the same interconnected ecosystem. These are the kinds of bots in the 10% that are thriving

Given the immense scope, quality, and impact of your work over the years, you manage to keep a pretty low profile. For instance, even long-time readers might not realize that you own UX Magazine.

Yeah. You know, for me, UX Magazine was never about making money and it certainly wasn’t ever part of a plan for self-promotion. My goal was to empower the experience design community to share ideas with a robust community of peers and arrive at a place where they could take an active part in what I saw as their larger goal—the goal of making sure technology wasn’t leaving people behind. Experience design to me is all about making it easier for people to get the most out of technology, and a lot of that boils down to the way people interface with machines. Over the years there have been many elegant and innovative designs that have made graphical user interfaces easier to use and that have connected new users with technology, but the reality is that GUIs can’t scale. Nobody wants to deal with 100 nested tabs and we have to do better. Luckily, as I’ve already mentioned, we now have access to an interface that everyone already knows how to use that is infinitely scalable: conversation. 

PQ: “Once conversational experiences stop sucking, adoption will skyrocket.”

The conversational AI platform marketplace is fraught with a lot of hype and unmet promises. How do you connect people with your unique vision?

I guess the easy answer to that is: “I wrote a book.” But really, anyone contending with conversational AI needs to be ready to do things in a a radically different fashion. This is especially true for organizations that are used to working with an assortment of vendors, each with their own closed system. The first thing to realize is that you cannot hyperautomate without some sort of shared communication layer connecting all of the people, systems, and things in your organization. This requires a platform that’s open and flexible. With that in mind, organizations should seek out vendors who are eager to create partnerships that will enable them to shape their own trajectories. Having a vendor team come in and build out a whole bunch of solutions can only get you so far. The best vendors in the conversational AI space are the ones that enable customers to build and evolve their own solutions.  That usually takes the form of co-creation, which is something that needs to run throughout any organization taking these efforts seriously.I like to say that AI is a team sport. Pairing solution designers with the people who best understand the tasks being automated unlocks a force-multiplying effect that can quickly put companies years ahead of their competitors. New software solutions can be created quickly, and then tested and improved upon without the limitations of traditional development cycles. It’s hard to overstate how powerful this paradigm can be.

Can you talk about the ways no-code technology fits into this paradigm?

No- and low-code creation tools make it so that the people building solutions don’t need any of the skills typically associated with a software developer. This can completely change the way companies hire solution designers. They don’t need to  look for unicorns who can write code while understanding the nuance and intricacies of design. Building with a code-free platform, the software is, in effect, the direct result of the design process. This frees companies up to hire people who are passionate about solving problems. One of our most successful lead solution designers at OneReach.ai came to us right out of college with a biology degree. It made no difference to us that she didn’t have a background in technology or design. We saw her as a crack problem solver, and in a few short years on our platform she’s probably designed more unique conversational experiences than entire teams at enterprise-level companies.

On a related note, I certainly don’t think it’s a coincidence that more than half of our designers are women. I was raised by strong women who got shit done—they were outcome driven, just like the designers on our team. I hesitate to make sweeping generalizations, but I’ve also seen that it can be easier for women to access empathy, which is such a big component of experience design.

You’ve mentioned the goal of technology not leaving people behind a few times, how broad is that initiative for you?

I view OneReach.ai, UX Magazine, and Age of Invisible Machines as three pillars supporting the shared goal of technology not leaving people behind. As I mentioned earlier, experience design practitioners are on the front lines of this struggle, and when they do their work well, the gap between people and machines narrows significantly. As an interface, conversational AI will have a massive impact on the ways people interact with machines. The simple fact that, in many scenarios, those interactions will take place without a screen should be of real interest to UX practitioners, and Age of Invisible Machines is definitely a book I hope many of them choose to read. Conversational AI and hyperautomation are inevitable forces that everyone will have to reckon with, and I want to arm anyone who’s interested with the information they need to take an active part in the maturation of these disruptive technologies. OneReach.ai is a tool I created so that more people will be able to build intelligent, conversational automations—which is really software creation—using code-free tools that can be leveraged without extensive training. Longtime readers of UX Mag may have noticed that OneReach.ai has appeared in some of the articles we publish, but there’s a reason behind that. Anything we share from anyone or any company, including my own, is published because it gives readers relevant, actionable insight. The bottom line is that we want more people taking part in the establishment and evolution of world-changing tech. 

You grew up under the direct tutelage of Canadian philosopher Marshall McLuhan. How did that shape your world view?

It was interesting because he was a friend of my mother’s who lived down the street, so he was more like my lovable, grumpy uncle. As I remember it, he really became fascinated with media and technology because of how much his wife liked to talk to her friends about the goings on in the neighborhood on the phone. His ideas were centered around the printed word, telephones, radio, television, and the kinds of computers that took up entire floors of buildings, so it would be interesting to know if he’d feel validated by where we’ve arrived with technology. I didn’t realize how renowned he was outside of my block until years later, but he instilled in me a deep reverence for the power of technology. 

One thing he said that I still think about all the time is this: “We shape our tools and thereafter our tools shape us.” I see evidence of this everywhere. When engineers at Facebook created algorithms with the positive outcome of getting the peak number of eyeballs on posts, they probably didn’t realize that evolved versions would imperil the mental health of an entire generation on Instagram, or that their news feed algorithm would undermine the mechanics of democracy. We are at a critical impasse right now. Conversational AI can quite literally help humans fix the world, but only if it’s used in thoughtful and responsible ways. This will be a tall order as companies the world over scramble to keep up, but I have hope that we can make technology work for humanity in profound and positive ways.

An abridged version of this interview originally appeared in Raconteur’s “AI for Business” report in the June 12th edition of The Sunday Times. Get the report here.

post authorJosh Tyson

Josh Tyson, Contributing Editor Josh Tyson is a Denver-based content specialist focused on UX, AI, and conversational design. He is the co-host of N9K, a podcast from the future, and stays balanced by drawing on the powers of family, skateboarding, yoga, and art.

Tweet
Share
Post
Share
Email
Print
Ideas In Brief

Related Articles

Learn ways user experience research can help build responsible AI applications.

How User Experience Research (UXR) Cam Help Build Responsible AI Applications
  • UXR reveals the myriad of intended and unintended effects of apps on people’s life — and help build more ethical AI
  • The author talks about:
    • How to assess the impact of ML-apps on users and meet AI ethical standards?
    • Ethics in AI
Share:How User Experience Research (UXR) Cam Help Build Responsible AI Applications
3 min read
How User Experience Research (UXR) Can Help Build Responsible AI Applications?

When less is usually — but not always — more

 

Balancing Cognitive Load and Discoverability
  • The author explains how reducing cognitive load helps user interfaces seem easier to use.
  • The article covers:
    • Information Architecture & Cognitive Load
    • Staggered vs Flat Information Hierarchies
    • The case for a flat hierarchy
    • The case for a staggered hierarchy
  • Balancing cognitive load and information density requires a lot of skill and deep knowledge of your customer.
Share:Balancing Cognitive Load and Discoverability
3 min read
Balancing-cognitive-load-and-discoverability

Tips on how to design for the blind.

Design for the Blind

The author explains how to design for people using Assistive Technologies and offers 5 tips.

Share:Design for the Blind
4 min read
Design-for-the-Blind

This website uses cookies to ensure you get the best experience on our website. Check our privacy policy and