One of science fiction’s most persistent promises is that, at some point, humans will be able to talk with machines just like we do with one another. Until very recently, “talking” with a machine mostly consisted of canned responses from impersonal machines and (often) frustration from us. But technological reality is quickly catching up with Hollywood magic, and it’s opening the doors to capable, versatile Artificial Intelligence (AI) that can converse with people in a natural way – and, dare we say, even exhibit a personality.
That’s where Amelia comes in.
Crafting software that successfully navigates the intricacies of human language is no easy task. It takes countless hours from a host of dedicated people. One of the pioneers helping to push the limit of conversational technology is IPsoft Cognitive Content and Conversational Experience Lead Monica Lander. With a background in communications and a fascination with technology’s potential to reimagine how we work, Monica is helping to evolve Amelia’s natural-language interactions and bring us ever closer to the sci-fi vision of a “talking machine.”
This is the first in our new series “The Real Women Behind Amelia” in which we profile some of our amazing colleagues who are helping write the next great chapter in technology, with Amelia front-and-center. Monica talks with us about the unique path which led her to working with Amelia, how she goes about crafting conversations in different environments, and how we can get more women to choose career paths in AI.
How would you describe your role at IPsoft?
My job is to think about how we can create the best experience for the end user through conversation, and to push forward improvements to keep evolving the natural language capabilities, and the humanization of Amelia.
What are you currently working on?
Currently, we’re developing conversational blueprints for common use cases across the banking and insurance industries.
The goal is to give Amelia industry-specific experience, including tasks, that she can handle right out of the box, so that businesses can immediately integrate her into their systems, build up her knowledge and refine the conversation.
What does the process of crafting conversations with Amelia look like?
First, I take a step back and think about — actually visualize — how a two-way conversational interaction will play out. It’s about being proactive and reactive.
To do that well, I need to understand the use case, the business requirements, the client’s goals and, most importantly, who is the user. How do they talk? How do they ask for things? What do they expect? What are their pain points? What do they truly care about?
Once we understand that, we can shape how conversations will be structured in order to complete a task, including planning for all the different things a user could say or ask about. Like screenwriters, we have to guide the story to its intended conclusion, while also planning for the twists and turns of the plot.
The ultimate goal for me is not to limit the intricacy and fluidity of real human conversation.
Are there specific things you think about when designing for mobile versus web versus voice?
Amelia’s web UI [user interface] is responsive, so when we build a conversation we think mobile-first. Planning the interaction for the smallest screen forces us to think though the most essential pieces of the conversation — things like word choice and UI elements. We try to make the conversations simple and keep Amelia’s responses relatively short, similar to how you would text with a friend. We also use UI features such as buttons, hint-text, and images to make the user experience easy, and enjoyable.
With voice, users have no visual guidance, so we need to take a different approach. It’s our job to set expectations up front and carefully guide the user through the conversation to the intended goal. We’re still in the early stages of realizing the dream of a “talking machine,” but I think recent advances in automated speech recognition are getting us closer. I think engaging with voice interfaces will become the new normal and the future of conversational UI will be a hybrid interface.
What do you think about when designing Amelia’s personality traits?
We want Amelia to adapt her responses to deliver personalized content and language depending on not only who the user is, but also on the company brand.
When we work with clients, we first ask to see if they have any voice and tone guidelines. This gives us a good idea of the attributes that make up the brand. If your brand was a person, what would she – or he – be like? Conservative or edgy? Witty or poised? Energetic or calm?
Answering these questions allows us to create a distinctive voice. It’s through this voice that we define how Amelia interacts and how we can give users a cohesive experience that is in-line with the brand.
What inspired you to want to work on AI?
I was a journalism major in college, but soon after graduating I went into marketing and corporate communications, before transitioning to digital content and strategy.
I’ve always been fascinated by language and have had to think through how to use language to connect with different audiences. When I started my career, it was relatively straightforward. Marketing and communications went mainly one-way – brochures, ads, e-mails, etc. Then technology made things more complex. All of the sudden we had to think about two-way interactions – social media, blog comment sections, text. My role naturally evolved into a more digital one, and I specifically sought out knowledge and opportunity to learn as much as I could about the new technology landscape.
When I was working on content strategy for IBM Watson, I realized that AI would have a substantial and transformative impact on our daily lives. That’s when I joined IPsoft. At IPsoft, I’ve had the opportunity to work directly on helping to develop Amelia. The potential to scale services, to eliminate mundane tasks so we can focus on more meaningful ones—and all through natural dialogue — is incredibly exciting.
How do you think the industry can get more women in STEM fields in general, and AI in particular?
I think interest in STEM fields starts at a young age, and needs nurturing through adolescence. I read a study that found that most girls view themselves as creative and are specifically looking for a career that can creates a positive impact in the world. I think girls in particular are somehow missing that connection with STEM jobs. As an industry, we can educate girls about the types of jobs that are out there and what their impact can be. Jobs in AI have the incredible potential to actively improve people’s lives and create positive change in the world – that’s something that a lot of girls, and boys, can help drive.
FIVE FAST FUN FACTS
Favorite depiction of AI from science fiction?
The AI from the movie Her. That movie is fascinating to me because of the interaction between Thor (human) and Samantha (OS), and the role that emotions play in their relationship.
Name a technology you’ve recently come in contact with that made you think, “Wow, I’m really living in the future.”
For me it’s VR (virtual reality). It has the potential to redefine how we interact with physical space.
Name a scientist or innovator whose contributions inspired you.
Noam Chomsky because of his work in linguistics and as a founder of cognitive science.
What’s something your co-workers would be surprised to learn about you?
English was my third language (Spanish and Romanian were first).
Your last meal would be?
French fries and an ice cream sundae.