Chatbots are becoming more critical to developers in their daily lives – from understanding how the technology operates, to creating better code. Developers tend to have a natural curiosity about bots and the tech behind it. Artificial intelligence tools exist to address emotional intelligence with chatbots in conversational interfaces.
Anamita Guha, product manager at IBM Watson, spoke about designing conversational interfaces at Spark the Change France 2018. InfoQ is covering this event with articles, summaries and Q&As.
InfoQ spoke with Guha about how chatbots can support developers in their daily work, how developers feel about chatbots, what it takes to design a chatbot that deals with emotions, and what the future will bring in conversational interfaces.
InfoQ: Where do chatbots fit into the developer’s platforms and tools, and how can they support developers in their daily work?
Anamita Guha: Chatbots can be used in a variety of ways to support developers: documentation (use it like an index); personal assistant (develop chatbots with a schedule to eliminate users’ daily routine); and ultimately, make lives easier by automating systems. Developers are all about efficiency and will hack together code motivated by making their lives and everyone’s around them easier. I know developers who have created bots to generate powerpoint status decks for executives. Cases like this help to free up their time to work on bigger problems.
Many use artificial intelligence (AI) and bots to automate all sorts of tasks and alerts like code deployments and testing. Additionally, when working in large, distributed teams, developers can see when an item is being pushed, merged, or edited by someone else, allowing more efficiency in development, collaboration and accuracy, and ultimately better code.
InfoQ: How do developers feel about chatbots?
Guha: Chatbots are seen to be emerging tech, and a lot of companies are in the hype: "We need a bot, and we need it now". Since 2015, the bot landscape has quadrupled, which means developers are realizing more than ever that chatbot development is important to learn as soon as possible.
Developers fall on a wide spectrum of emotion when it comes to chatbots, depending on who the developer is and what the bot is used for. In general, most developers want to stay ahead of the curve when it comes to development so they can stay relevant in such an evolving industry. This fosters a natural curiosity about bots and the tech behind it. In terms of using bots, I’ve found developers will use bots only if it is useful to them. This includes web-based bots, as well as voicebots like Alexa and Google Home.
From start-ups to SMBs to enterprises, developers are facing similar challenges; it is hard to always find a business justification to the executive team on why a bot might be necessary, or even required. Simultaneously, the leadership team is trying to balance their need to evolve as an organization by leveraging new technologies like bots.
For these reasons and more, we are now at an inflection point. For more than a decade since the Apple store’s inception, everyone wanted to build an app, but now, people want to build a bot. Today, bots are even becoming the gateway for kids to get into coding and the first step in development – and ultimately to get involved in technology. Bots are becoming more and more pervasive and multi-modal.
InfoQ: What does it take to design a chatbot?
Guha: I approach chatbots and their design the same way I do a product, by asking the following questions: what problem does it solve, who is our target user, and why should they care? To answer the questions successfully, it is imperative to first understand the use case, where it will live, and the demographic of the users. This is because it influences the conversation design. For example, when designing a conversational chatbot on Facebook used by teens, you could use emojis, whereas an Alexa skill for banking transactions would use direct and simple verbiage.
It’s also important to note that bots, like the evolution of human communication, are evolving. Learn from how users are interacting with your bot, and iterate!
InfoQ: How can you make a chatbot that knows how to deal with feelings?
Guha: Various AI tools can help you address emotional intelligence. For instance, the empathy suite at IBM helps users see how someone is feeling based on how they are typing. Different words can give a certain tone or vibe, and one can respond accordingly.
This will also vary depending on the medium the bot manifests in. If it is a text-based bot, consider word choice and sentence structure – or a voice bot: tone, voice, and inflection. Additionally, a virtual assistant or entity bot in AR/VR is wired with eye tracking and can pick up on body language, so be sure to research technologies to help you find the right tool that will empower to build the emotionally intelligent bot.
InfoQ: What will the future bring us in conversational interfaces for platforms and tools?
Guha: In the future, I think everything will be a conversational interface. For example, you will be able to instruct anything, from your phone to your watch to your car, to perform an activity, and they will all have a working memory of the task. We already live in a world where transactions are done through biometrics like eye tracking or fingerprints. As these futuristic notions come together, they will ultimately allow developers and technologists who are building this future to obtain a larger digital footprint on someone and eventually personalize how one lives.
InfoQ: IBM started the initiative Call for Code which aims to support innovation and technology for good. Can you elaborate?
Guha: Recently, there has been a wave of AI and intelligent technologies that seem to be driving a greater emphasis on the use of technology to create social good. While the tech industry has somewhat of a reputation for being self-serving and capitalistic, I think everyone’s cognizant that they want to give back — even if they don’t necessarily know how.
One way IBM is addressing this is through the Call for Code. Natural disasters have been growing in severity and frequency over the past decade. While we cannot stop them from happening, we can use technology to lessen their impact. We’re encouraging developers worldwide to use their skills and our technology to help solve the problems natural disasters create.