• Michael Wright


Some choice-based interfaces pre-date the home computer. “Choose your own Adventure” books were popular in the early 80’s, and they involve a branching path format in order to move between sections of text, but the reader creates a character as in a role-playing game. Choices made in the books determine the outcomes of actions and created potential for different experiences. Many more examples can be found in gaming, as the demand for personalised experiences becomes less of a novelty, and more of an expectation.

The two main types of conversational interfaces these days, are bots and voice activated assistants. They can help train you, they can help you order the perfect Taco, resolve your calendar woes, and greatly enhance your reading experiences, as showcased in the brilliant Typeform article featured further down in this article.

Conversational interfaces work well everywhere, including smartphones, desktops, smartwatches, and even devices without screens. They can integrate with web services like Facebook, WeChat or Snapchat, or run just in a text message window. They can be improved gradually, over time without requiring completely new versions. Conversational interfaces can also represent the full features of a service or app without having to render menus or clickable icons, but to focus on guided user discovery instead. A well built CUI will delight the user by taking the weight off them and guide their choices.

With text recognition and processing improving, we are on a path that could make interaction with digital services more intuitive, accessible and more efficient. They will allow humans to talk to computers in a way that puts the onus on the software, not the user, to figure out how things work.

Conversational interfaces are, of course, also highly interactive. Higher interactivity enhances recognition as well as recall memory of interactive content. Not only that but studies conclude that interactivity promotes a higher emotional response and greater enjoyment.


For consuming content in a totally interactive, non-linear way check out this great article by Typeform titled “Technology Imitates Art, The Rise of the Conversational Interface”

Pull Google’s Allo into any conversation just by mentioning (it) and get things done.

TacoBot doesn’t just help you sort your choices of fillings, but can even answer the age-old conundrum about the chicken crossing the road.

So how does one design a conversational interface?

I’ve been playing around with bot-building apps like Motion.ai for a while now, and I’ve learned to be as clear as possible and present users with relevant options.

Avoid open-ended questions, and instead suggest answers. Play back to clarify the users thought or instruction. Make the conversation feel as social and natural as possible, not just functional. Just like in comedy, timing is everything, and can build anticipation, so use the power of the pause.

Where I work, over at Grab, we’ve already introduced “Fishbot” who is a Slack-based Chatbot who can provide answers on day-do-day HR questions to enhance the on-boarding experience but I feel we’ve just scratched the surface.

How will conversational interfaces impact recruiting?

One of the biggest challenges facing recruiters is improving the demoralising candidate experience. If one of the major issues is the lack of attention afforded to applicants, based on a limited capacity of human recruiters, can conversational interfaces bestow much needed attention?

Given the vast volume of unfilled positions, will they help us get better at selecting talent? Can they tease out the information needed to find the hidden gems who don’t qualify on paper for the a specific job?

If identifying and selecting talent become more automated, will recruiters be made redundant as companies strip costs out of the bottom line? Or will we be empowered superheroes, working in perfect harmony with our neural networks?

How will you feel about being ‘interviewed’ by a conversational computer?

Whilst Machine Learning for Image Recognition holds a lot of promise, like for instance enabling T-Rex to come alive in downtown Singapore. it’s also possibly going to lend new credibility to old notions around physiognomy (bias and inference from pictures) covered on Medium in this article by Blaise, the leader of Google’s Machine Intelligence Group, and this controversial research paper which implies that can predict the likelihood that a person is a convicted criminal with nearly 90% accuracy using nothing but a driver’s license-style face photo.

How will you feel if a computer is used pass judgement on your suitability for a job based on your Linkedin profile pic alone?

As Textio’s CTO, Jenson Harris points out, although many conversational interfaces might have ‘personality’, they could actually be a “Hipster facade around the same basic command-line interfaces we left behind in the 1980’s”

Like Jenson, I’m think bots are cute, but I’m more excited by the application of artificial intelligence to vastly improve internal business processes. He quips:

“While being able to talk to your CRM is cool, having a sales platform that accurately predicts the 100 opportunities you can close this quarter is worth breaking the bank for. Having a cute avatar answer your customer support chats seems nice enough, but predicting ahead of time what areas of your product will get support requests so that you can fix them before customers suffer is pure gold”

I’m personally most excited by the idea that AI and deep learning algorithms can help us to predict the future and inform us how to alter it for the better, but certainly not to grant control to them in totality.

#chatbots #AI #conversationalinterface #CUI #tacobot #motionai