Chatbots

Introduction to Bots

What are chatbots and where did they come from? Early roots in psychology and an ironic twist behind the creation of the first chatbot paved the way for modern chatbots. Today, chatbots have become a popular customer service solution able to facilitate sales, complete online tasks, and more.

April 3, 2018
Introduction to Bots

Roots in psychology

It might feel like bots are just arriving on the scene, but bot technology has been around since 1966. In fact, German-American computer scientist Joseph Weizenbaum invented what is generally accepted to be the first-known bot (known as ELIZA) to mimic conversational speech patterns more than 50 years ago. He created ELIZA to prove that machines could not replace human conversation. In a twist of irony, Weizenbaum’s ELIZA proved him wrong—and paved the path for bots.

ELIZA was designed in the pattern of Rogerian therapy, also known as “person-centered therapy.” Psychologist Carl Rogers posited that each human has the power to find the best solutions to their own problems. Rogerian therapists believe that patients come to their own conclusions about issues and solve their own problems with the emotional support of a therapist. Ultimately, the patient controls the majority of the conversation. The therapist listens empathically and without judgment. The therapist’s only role is to encourage and support the patient—with no interruptions or interferences.

What a perfect role for a bot! A bot is not emotionally invested, and it can be programmed to respond with patience. Additionally, a bot “listens” but doesn’t offer solutions. It provides the factual answers it was programmed to deliver. If you tell a bot that you feel sad, it may simply parrot, “Why do you feel sad?” And you might feel like it’s really listening.

Was ELIZA’s simple mimic therapy effective? Absolutely. Contrary to Weizenbaum’s expectations, people were enamored with ELIZA. Her talk therapy made many “patients” feel heard and understood—and she even made them feel better. ELIZA convinced some people that she was intelligent or that she may have even been secretly controlled by a human.

ELIZA’s success made psychologists even more interested in the potential of bots. In 1971, psychiatrist Kenneth Colby invented PARRY. This bot simulated the language response patterns of a paranoid schizophrenic. Colby used the robot for teaching his psychology students, since it modeled schizophrenic behavior perfectly. And, in one experiment, ELIZA took PARRY on as a “patient” to see what Rogerian therapy could offer.

Read a transcript of the conversation between ELIZA and PARRY.

After ELIZA

A bot provides users with programmed answers to questions that it recognizes. If programmed to do so, a bot can theoretically provide an infinite amount of information. The question for developers might have originally been, “What can we program this bot to do?” But it soon became, “What can’t a bot do?”

ELIZA and PARRY rapidly gained worldwide fame. Soon, many computer developers were wondering about the possibilities presented by bots. As long as the bot followed the logic of Rogerian psychology (listening and providing information), couldn’t there be other uses?

Many iterations of bots were created throughout the decades: Jabberwacky, 1988; Dr. Sbaitso, 1992; Artificial Linguistic Internet Computer Entity (A.L.I.C.E.), 1995; and SmarterChild, 2001. The goal was uniform—each bot needed to be practical and to instantaneously provide a set (or sets) of information.

One of the biggest successes ultimately came from the technology within smartphones. In 2010, Apple launched Siri for iOS. Siri was the first multi-functional, voice-commanded bot. She paved the way for intelligent personal home assistants, like Amazon’s Alexa, which have become household names around the globe.

Bots have even gone back to their roots. Medical and mental health care providers are experimenting with bots for diagnosis and ongoing care. A study from Juniper Research suggests that by 2022, 75-90% of healthcare queries will be managed by chatbots.

If you build it, they will chat

Nowadays, the central use of bots is even more comprehensive. Today’s bots aren’t tied to personal assistants, and they’re not just mimicking therapy. They are migrating to messaging apps and websites. What started as talk therapy is flourishing in text and speech. And Weizenbaum’s little thought experiment is now a standard business practice.

Bots have become a popular customer service solution for private enterprise. They’re constantly available to customers, maintain the voice of the brand, and even work for free—after you build them, of course. They can be used to route communications more effectively to customer service representatives, or they can act—all on their own—as the rep.

They can even be used for direct sales. For instance, if a user is shopping for a new pair of shoes, a bot can ask questions about the shopper’s shoe size, model preference, style, and price range. The bot can provide the shopper with exactly the product they’re searching for in a way that feels tailored to that user’s needs. Consequently, bots are turning conversations into conversions.

So where will this trend go next? These days, industry experts are speculating about how enterprise bots will help large businesses automate workflow across platforms and channels. Bot adoption is off to a strong start: 38 percent of enterprises already use them, and an estimated 62 percent of them will do so by the end of 2018.

Small businesses are using bots, too, because the majority of customers have expressed the desire to use a bot when given the option. Additionally, for the first time in history, people are using messaging apps more than they’re using social networks for communication.

What’s the adage? If you build it, they will come. In today’s business environment, if you build it, they will chat.