Conversational Interfaces as an Emergent Property
Last week I had the pleasure of giving a guest lecture at the University of Limerick on the subject of Internationalising Digital Services companies. I was happy to share some of the challenges that a startup faces, the trade off’s it needs to make, and to emphasise that while there are no silver bullets, there are patterns we can all learn from.
It wasn’t until I was actually at this slide, about to speak to it, that a new thought occurred to me:
Textual Communications: not only are we moving from SMS to messengers but these messengers are now private social networks. For myself I also don’t send personal email anymore. Messengers ate my text interactions. Messengers are also where I scroll back to find, confirm, remember. It’s taking some of my cognitive load. It’s where I leave my memory when I go to the shop. That’s why it also needs to be searchable, and it needs to get predictive. The new Google Assistant on it’s new Pixel phone does just this in that it even helps you keep the conversation going with someone by suggesting topics.
Voice Interaction: Alexa is a niche voice activated service that we mostly use to turn the music on in our homes, and the lights off. But we are being trained to ask virtual assistants to do things for us. Over 20% of mobile searches are now voice activated, and these are three times more likely to be for “local search”. It makes sense if you ask your iphone “what’s a good sushi place around here”. So while I may have the wobblies around being faced with a “voice IVR” my kids will grow up in an Alexa world where they are speaking into thin air the whole time anyway.
Visual Interaction: Snapchat encourages you to group photo’s into a sequence that it calls stories. But it’s not the sharp memorable image that is valued, but the narrative that is placed over or implied by the sequence. Great day in Dublin. Market. Sun. Lunch. Drinks at night. I find myself not sharing individual photo’s as much, but sequences, from which there is an implied narrative.
What struck me so forcibly was that we have been living with these communication channels for years. Text. Voice. Pictures. What’s different is that they are doing subtle things differently in ways that implies continuity, a narrative. The emergent property of these networks is the invisible assisted interaction.
We are prompted to comment on pictures, to name the sequence, to overlay the meaning. We are given new interaction affordances that encourage us to group actions, share the thread, or to add and subtracting objects in a way that feels different these days.
The way a messenger encourages us to collect, name, and tag gives us the raw materials that move us towards the experience of “conversational”. A.I. (artificial intelligence) is being applied to the text, to the words you say, to the photo’s you share (for instance, visual recognition systems that allow google to show you “all my pictures with a dog on the beach from last year”).
Intelligent Assistants are clearly the next step in moving us all down this conversational path. Artificial Intelligence will be important at the conversational platform level. The question will be how much of it comes “packaged as features within the products you buy from vendors”, how much of it is a general capability you have access to from cloud based API’s, and how much of this you are going to have to build and wrangle with for yourself?
While all this may sound “a bit out there” let me assure you companies are already in action mode today to bring these capabilities into the enterprise B2C conversations.