What Talkabot 2016 tells us about the present and future of computer conversation

By Sarah Wulfeck

Thousands of the best and most prominent chatbot writers, developers and messaging platform reps attended one of the first conferences centered on chatbots — Talkabot — hosted by Howdy.ai in Austin, TX in late September.

I went to introduce the bot world to our computer conversation authoring tool, PullString Author, showcase the projects and experiences we’ve authored, and to learn what about what others are doing in this space.

Here are some of the highlights, and what I took away, as a veteran writer/director of computer conversational experiences.

Me, on the left, letting my colleague Lucas Ives take a turn on the mike.

The bar has been raised for content quality.

I was pleased to see that many Talkabot attendees and speakers are taking the creative capabilities of computer conversation seriously; acknowledging the new art form requires authorship from creative professionals.

“Giving your bot a soul is not a technical problem, it’s an editorial problem.” — @roberthoffer #talkabot
— Ben Straub (@benstraub) September 29, 2016

Vivian Rosenthal, CEO of message-based marketing service, Snaps, said the company spends about 90 percent of their efforts on script writing, as opposed to tech and programming. Facebook Messenger’s team stressed the importance of storyboarding chatbot experiences, and Kik’s CEO, Ted Livingston, confirmed the chat service’s emphasis on experience.

This is an encouraging trend, and we’ve been seeing it in the broader public over the past year, too.

High-profile, premium content bots like Kik’s CNN bot, Microsoft’s Chinese chatbot phenomena, Xiaoice, and the Call of Duty: Infinite Warfare bot have been published to the delight of the public, getting people more excited about the possibilities of conversational interactions. Last week’s news of Google hiring creatives from Pixar and The Onion for Assistant is further confirmation of this trend.

For a deep dive into chatbot writing and scripting, check out our ebook “Create Convincing Computer Conversations,” which includes a thorough walkthrough of chatbot character development, writing and scripting advice, and writing prompts to sharpen your skills.

The industry is still divided on AI and scripting, but hybrid approaches are gaining a foothold.

Much time was devoted to the merits of artificial intelligence in bots. Historically, the debate between primarily AI-driven bots and scripted bots has implied a false choice between one or the other. Now, more developers are combining the capabilities of AI with the warmth and delight of good writing.

At this year’s conference, Snaps, AdmitHub, Voxable and Kik all emphasized a hybrid approach where technology serves as a computer conversation enabler, not a driver.

“Use AI to augment your bots but don’t make it the point of it.” — Kik CEO, Ted Livingston

This doesn’t mean AI’s contribution to good computer conversation is negligible, though. As Alessandro Vitale, Founder of Optimist AI., describes in his presentation, advanced natural language processing (NLP) is essential for a user experience that’s smoother, more responsive and values user discovery.

However, despite its massive potential, we’re still in the early stages of NLP development. The technology hasn’t advanced enough for developers to rely on it entirely — if they want their experiences to be believable.

Cross-chat-platform support is becoming a must-have, but platform capabilities are increasingly distinctive.

Among the clearest desire on the part of bot developers was cross-platform publishing — this year developers clearly want bots made for Facebook Messenger to work in Slack.

Message.io announced they were working on cross-platform support, a scripting language for “global” chatbots, and a conversion tool. But while cross-platform capability is a clear improvement, that doesn’t mean all bots have to work on all platforms.

As Slack put it in their post-Talkabot analysis, differing capabilities and audiences among messaging services means that even if a bot functions on all platforms, it won’t necessarily work.

“Sure, it would be nice to write a simple bot that worked on both Kik and Slack without having to redo code, but an app used by teens on Kik would likely be much different than one used my middle-aged adults in a business context on Slack.”

I typically recommend brands looking to create bots on these platforms to think first about their customers and desired audiences, decide which platform they prefer, and publish their bot there. Both the voice and characterization of the bot, as well as the platform it’s published to, matter to its success.

So… the future of chatbots, and the art form of computer conversation. What’s next?

I’m highly interested in seeing what newcomers — and old pros — make as chatbots are more creatively developed, and the field’s impact on users is deepened.

We’re at the beginning of a revolution, as advanced, AI-driven capabilities and expressive development collide. I’m looking forward to seeing where and how these capabilities and developments grow, how to simplify the complexities of authorship, and to leverage these technological advancements for even more engaging computer conversation experiences.

For more, check out the Talkabot presentation I gave alongside and PullString’s ‎Head of Conversation Engineering, Lucas Ives: “The Art And Science of Computer Conversation.”