Over the last year or so, I have been spending a lot of time thinking about the relationship between our mobile technologies and ourselves to understand how can technology mediated services do even a better job at satisfying our individual information needs and save time creating, processing, sharing and using information to get things done. It’s things like helping with taking care of parents and their medical needs, organizing meetings and tasks to negotiate mergers and acquisitions, and developing an enterprise portal integrating expert knowledge in Chatbots to help researchers find relevant IT resources. As a strategy, I see a large unmet need in figuring out cost effective ways to create a individualized experience, but believe that despite some noted challenges, the tech industry is breaking through with artificial intelligence driven platforms that may work.
One of the emerging tech trends that has captured my attention are bots in general, and particularly Chatbots. Bots are software applications that run automated tasks. Chatbots are a type of bot that “is an artificially intelligent software program that uses natural language processing to simulate intelligent conversation with end users via auditory or textual methods” (https://playground.pandorabots.com/en/). Chatbots serve as virtual assistants, conversational agent and virtual customer service agent.
With Google, Microsoft and Facebook making industry announcements in the last few weeks about their long-term Chatbot strategy, industry analysts are declaring that 2016 is the dawn of the “bot economy”.
On one hand, bots are not new and are in broad use today. Joseph Weizenbaum, a MIT professor, created Eliza, the first Chatbot, in 1964-1965. Professor Weizenbaum wrote Eliza using artificial intelligence programming to allow a person to have a conversation using regular every day English with a computer. Eliza mimicked human conversation by modeling active listening techniques that many psychiatrists use to communicate with their patients. Eliza would reframe the user’s statements as questions. Eliza was one of the first attempts to create the illusion of human-to-human interactions using a computer program in human-machine interactions. In Professor Weizenbaum’s obituary, the New York Times reports, “in fact, the responsiveness of the conversation was an illusion, because Eliza was programmed simply to respond to certain key words and phrases. That would lead to wild non sequiturs and bizarre detours, but Mr. Weizenbaum later said that he was stunned to discover that his students and others became deeply engrossed in conversations with the program, occasionally revealing intimate personal details” (Markoff 2008, http://www.nytimes.com/2008/03/13/world/europe/13weizenbaum.html?_r=0). Professor Weizenbaum in later years used Eliza to raise societal questions about what is the appropriate type of relationship between humans and computers.
Prof. Joseph Weizenbaum in Berlin, Germany (Source: https://commons.wikimedia.org/wiki/File:Joseph_Weizenbaum.jpg)
Jumping 50 years ahead to today, Chatbots and other bot varieties drive much of the traffic on the Internet. Imperva Incapsula reported back in 2012 that the majority of traffic on the Internet was computer generated through bots. 51% of website traffic was non-human traffic. 20% of the traffic was comprised of good bots for search engines and so on. “Good bots are the worker bees of the Internet that assist its evolution and growth. Their owners are legitimate businesses who use bots to assist with automated tasks, including data collection and website scanning (https://www.incapsula.com/blog/bot-traffic-report-2015.html). The remaining 31% of the traffic was comprised of bad bots used in hacking tools, scrapers, comment spammers and other spying algorithms. “Bad bots, on the other hand, are the malicious intruders that swarm the Internet and leave a trail of hacked websites and downed services. Their masters are the bad actors of the cyber-security world, from career hackers to script kiddies. In their hands, bots are used to automate spam campaigns, spy on competitors, launch denial of service (DDoS) attacks or execute vulnerability scans to compromise websites on a large scale” (https://www.incapsula.com/blog/what-google-doesnt-show-you-31-of-website-traffic-can-harm-your-business.html). For example, bad bots include those that are used in click farms to inflate the value of social media accounts (Clark, D. B. (2015). The BOT Bubble. (Cover story). New Republic, 246(4), 32-41). In the most recent 2015 report, Imperva Incapsula reports that 51% of the web traffic is human generated, which is the first time in five years (https://www.incapsula.com/blog/bot-traffic-report-2015.html).
Leading tech giants Apple, Google, Facebook and Microsoft are betting on Chatbot platforms to drive commerce and new business models. With recent announcements, the tech industry believes the world will soon move away from apps—where Apple and Google rule—into a phase dominated by chats with bots (http://www.economist.com/news/business-and-finance/21696477-market-apps-maturing-now-one-text-based-services-or-chatbots-looks-poised).
Since 2011, Apple, of course, uses Siri in its iPhone, iPad, iTV, CarPlay and iWatch products to provide each user a personal assistant. Siri uses a natural language processing interface and artificial intelligence algorithms to allow the user to ask questions and issue commands.
Google CEO Sundar Pichai announced during the Google I/O keynote on May 18, 2016 that it is releasing Google Assistant to be an automated messaging helper. “We think of this assistant as an ambient experience that extends across devices,” Google CEO Sundar Pichai said. “It’ll be on their phones, the devices they wear, in their cars, and in their homes.” Google will offer this through and AI driven chatbot called Allo and a second AI driven product called Google Home. According to Google, “Allo is a smart messaging app that makes your conversations easier and more expressive. And with deeply integrated machine learning, Allo has smart features to keep your conversations flowing and help you get things done. Allo has Smart Reply built in (similar to Inbox), so you can respond to messages without typing a single word. Smart Reply learns over time and will show suggestions that are in your style. Allo also features the Google assistant, bringing the richness of Google directly into your chats—helping you find information, get things done, and have fun. You can chat one-on-one with the assistant, or call on Google in a group chat with friends” (https://googleblog.blogspot.com/2016/05/allo-duo-apps-messaging-video.html).
In mid-April, Facebook announced that it would deepen its use of Chatbots through its Messenger product. Facebook shared that “We’re excited to introduce bots for the Messenger Platform. Bots can provide anything from automated subscription content like weather and traffic updates, to customized communications like receipts, shipping notifications, and live automated messages all by interacting directly with the people who want to get them. The Messenger Send / Receive API will support not only sending and receiving text, but also images and interactive rich bubbles containing multiple calls-to-action. We’ve also built powerful discovery tools such as plugins for websites, usernames and Messenger Codes and a prominent search surface in Messenger. Last but certainly not least, the availability of Wit.ai’s Bot Engine will enable developers to build more complex bots that can interpret intent from natural language, and continuously learn to get better over time” (http://newsroom.fb.com/news/2016/04/messenger-platform-at-f8/). In addition, Facebook CEO Mark Zuckerberg explained, “Consumers already use Messenger to communicate with businesses. More than 1 billion messages are sent between businesses and users via the app, Facebook says. Messenger is already used to request a lift from Uber or Lyft, purchase tickets for concerts and other events, make a payment and more. The addition of Chatbots will make those connections deeper, with better services” (BRIAN, D. (2016, April 12). Facebook CEO Zuckerberg On Bots, Live Video And Global Access. Investors Business Daily. p.1).
Microsoft has developed two AI Chatbots for interaction with users in social media applications with mixed results. In China, Microsoft released the Xiaolce Chatbot and has generated over 40 million conversations. Microsoft released a second Chatbot, Tay, to experiment with a Chatbot tailored to provide services to a more diverse user base, particularly focused to talk like a 19 year old teen to users between 18 to 24 year olds in the US for entertainment purposes on live on Twitter, GroupMe, and Kik (Microsoft blog). Tay stands for “thinking about you”. As Microsoft writes “The logical place for us to engage with a massive group of users was Twitter. Unfortunately, in the first 24 hours of coming online, a coordinated attack by a subset of people exploited a vulnerability in Tay. Although we had prepared for many types of abuses of the system, we had made a critical oversight for this specific attack. As a result, Tay tweeted wildly inappropriate and reprehensible words and images” (http://blogs.microsoft.com/blog/2016/03/25/learning-tays-introduction/#sm.0000izjzez1ck0ej9x08yacj1aqix).
For a good list of Chatbots in action, please see https://www.chatbots.org. Here is an example of Chatbot services making their way to the marketplace.
CHATBOT
DESCRIPTION
Assist
An AI driven personal assistant to help with many everyday tasks https://www.assi.st
Chatfuel
A Chatbot builder with AI navigation for Facebook’s Messenger https://chatfuel.com
Digit
Digit is a personal assistant that helps you to save money based on analyzing your spending habits monthly. https://digit.co/about/how-it-works
Elbot
Artificial Solutions is the leading specialist in Natural Language Interaction (NLI), a form of Artificial Intelligence that allows people to talk to applications and electronic devices in free-format, natural language, using speech, text, touch or gesture. http://www.elbot.com/chatterbot-elbot/
MeeKan
A scheduling assistant robot for Slack of Hipchat https://meekan.com
Operator
A personal shopping assistant https://www.operator.com
Pana
Pana arranges your travel based your preferences https://pana.com
Troops
A Chatbot (slackbot) for work to help you use Salesforce https://troops.ai
X .ai
A scheduling assistant https://x.ai
Written by Jon Gant, LG CNS Blog’s Regular Contributor