What we learned at ChatBotConf 2016 in Vienna
We were in Vienna at the first European chatbot conference. More than 400 participants and 19 speakers met on 14.10.2016 and taught us a lot. A brief insight into the five most important learnings from the conference can be found here in this article.
Slack
Amir Shevat, Head of Developer Relations at Slack, sees bots in their ecosystem as a vital part of increasing productivity. With easy-to-program bots for Slack, you can bring more and more functionality of other tools used by the team into one interface. Instead of going to the corresponding calendar application for an appointment, the same functionality can also be operated with simple commands from Slack.
IBM
Christoph Auer-Welsbach from the IBM Watson team sees a triumph of chatbots in the service segment as very likely. According to studies by Gartner In the near future, a large part of all communication and interaction between customer and company will take place not only digitally but above all without human intervention.
“By 2020, customers will manage 85% of their relationship with the enterprise without interacting with a human.”
In order to live up to this responsibility, IBM relies on Watson and explains the understanding of human creativity as a goal.
microsoft
Toby Bradshaw from Microsoft's bot team gives important advice along the way: “Design for confidence.” The aim of a conversation and the prerequisite for providing the user with correct answers is to be sure of what the user wants. Instead of responding to the most likely need and hoping that's already true, Bradshaw advises simply asking, “Hey, I think you want X, is that correct?”
Meekan
When developing a chatbot, it is better to focus on one topic. Because if you don't sit on a heap of data and users, it will be difficult to cover everything. A chatbot should first become really good at a subject area. Be it weather or events. If users ask the bot things that it does not cover, it is better to dismiss this playfully but not answer the request half-heartedly.
Google's own Skynet — better known as Knowledge Graph — is best expressed in figures. More than two billion entities spread over 38,000 types and 54 billion facts about them now help the Google Assistant find answers to almost all questions.
The enormous advantage that Google has over the years of processing search queries and Google Voice data on just about every topic enables a sophisticated voice assistant such as the Google Assistant.
A scenario like in the movie “Her” isn't that far away anymore — just hopefully a bit less creepy.
The conference helped us to better understand the relatively fresh topic of chatbots. We met a lot of interesting people and the talks were all informative. If the conference takes place again next year, we will definitely be there.
UPDATE: We were there - 6 takeaways from Chatbotconf 2017
We would like to take this opportunity to thank the organizer orat.io for the very interesting conference.
