Though it may seem that engaging with AI like Amazon’s Alexa, Apple’s Siri, and Google’s Assistant is a relatively new phenomenon, it’s anything but. The basis for the technology and evolution of chatbots date back close to seventy years.

Before we recap their long and illustrious history, let’s get an understanding of what a chatbot is and what purpose they serve.

What are Chatbots?

It’s all right there in the name.

Chatbots are basically an AI program designed to engage or chat with humans through text or audio-based means.

Early bots exclusively used text-based communication and could then only converse in a rigid question-answer-question-answer format. As technologies evolved so too has the sophistication of the conversations.

The ultimate goal of this type of AI is to successfully mimic the way humans converse.

In fact, much of the evolution of chatbots stems directly from attempting to create a chatbot whose speech patterns and conversational mannerisms are entirely indistinguishable from that of another human.

It that all sounds a little sci-fi, just ask Apple’s Siri or Amazon’s Alexa a question to see how far the AI has come.

Chatbots, A History

The idea behind chatbots didn’t actually start with a bot, but instead a theory.

In the early 1950s, Alan Turing put forth the idea that at some point AI machine intelligence would be equal to that of a human. The “Turing test” postulated that a human would one day be unable to distinguish between the speech of a fellow human and that of a machine.

Ultimately, the theories Mr. Turing laid out in his 1950 paper, “Computing Machinery and Intelligence,” would lay the groundwork for the development and evolution of the chatbot.

Nearly twenty years would pass until the first AI iterations would start to test his theory.

1966 – ELIZA

An MIT creation from their AI lab in the mid-60s, ELIZA was intended to match patterns of speech thereby suggesting an understanding of human conversation. A major breakthrough in the development of early AI, ELIZA remained limited by the inability to conceptualize and place events into context.

ELIZA was one of the first chatbots to attempt the Turing test.

1972 – PARRY

Six years later, Stanford University developed PARRY, a design made to reflect that of a human with paranoid schizophrenia. PARRY possessed a far more conversational structure than that of response-driven ELIZA.

It’s worth noting that PARRY and ELIZA “met” on several occasions with ELIZA service as the therapist to PARRY’s patient. The most widely know of these interactions happened in 1972 over ARPANET, an early precursor to the modern-day Internet.

1988 – Jabberwacky

Originally developed in the early 80s, Jabberwacky was an early form of AI developed from human interaction. The specific goal of Jabberwacky, as noted by its British creators was to “simulate natural human chat in an interesting, entertaining and humorous manner.”

One of the significant things about Jabberwacky was its voice-operated approach to conversation versus simple text.

1992 – Dr. Sbaitso

Known widely for their SoundBlaster sounds cards which proved the standard for early PC audio, Creative Labs also developed AI therapist Dr. Sbaitso.

The program itself was fully voice-operated and more of a means to highlight the use of digital voice (and it sounded 100% digitized) more so than any groundbreaking AI conversational techniques.

1995 – A.L.I.C.E.

Genuine chatbot advancement came in 1995 in the form of Artificial Linguistic Internet Computer Entity otherwise known as Alice or Alicebot. Conversations with Alice were truly unique as the AI employed a style utilizing heuristic (or problem-solving) pattern matching.

Although unable to pass the Turing test, Alice simulates the dynamics of having a conversation with a real human over the Internet. Interestingly enough, the idea for the 2013 movie Her was born from the film director, Spike Jonze, using the Alicebot program a decade prior.

2001 – SmarterChild

Considered the bot that laid the groundwork for Apple’s Siri, SmarterChild was an intelligent, instant messaging bot produced for the SMS networks of AOL and MSN Messenger (now Windows Live Messenger).

Although geared more for entertainment, SmarterChild possessed capabilities allowing it to access data and carry on personalized conversations.

2006 – Watson

Once upon a time IBM’s ubiquitous Watson AI was limited to the role of Jeopardy game show contestant (albeit a reasonably successful one, besting two former champions).

The system grew beyond its fundamental underpinnings of answering questions in natural language and now serves as a data cruncher for a wide range of industries.

Considering its ability to process large amounts of information and reveal insights using natural language, Watson proves ideal for a number of uses. To date, utilization of the AI includes healthcare, weather forecasting, diamond buying, and customer service, among many others.

2010 – Siri

Apple’s Siri kicked off the AI digital assistant craze of the past decade, which has seen simple question and answer sessions with our phone morph into directives aimed at making our lives more comfortable and easier to manage through the Internet of Things (IoT).

Much like SmarterChild paved the way for Siri, Siri has proven the trailblazer for the chatbot AI we use today. Integrated directly into Apple’s OS across all of its platforms, the intelligent assistant has lost ground to more recent iterations but remains the baseline for all current AIs.

Although its abilities to converse and perform specific tasks were rightly heralded, arguably the most significant advance Siri brought to the AI space was a personable and accessible style.

2012 / 2016 – Google Now and Google Assistant

Developed for Google’s mobile search app, Google Now serves as a predictive search AI working to anticipate needs versus just responding to inquiries, although it does that too.

As the predictive capabilities of Now evolved, pulling its data directly from a user’s search habits and general activity, it became more intuitive and proactive with the information it delivered.

Now still remains a part of the Google functionality, though its branding has been phased out.

In 2016, Google launched a new AI driven digital assistant, Google Assistant. A definite enhancement over the Now functionally, Assistant allows two-way conversations and integrates with mobile and smart home devices.

2015 – Alexa

Siri may have been the catalyst, but Amazon’s hugely popular Alexa is the current standard bearer for digital assistant AI.

Fully integrated into Amazon’s line of smart devices like the Echo, Dot, and Show, Alexa also proves the popular choice in controlling a number of third-party devices that make up the Internet of Things. Developers can build in functionality or skills for Alexa, which results in an increase of the AI’s intelligence.

What might be the most compelling, and indeed, consumeristic aspect of Alexa is the ability to place voice orders through Amazon, including groceries. Of course, an Amazon account and supported device are necessary, but along with IoT functions, it represents chatbot AI managing more and more of users daily tasks.

2015 – Cortana

Microsoft’s entry in the AI digital assistant space, Cortana did not move the needle of chatbot AI forward in any appreciable manner upon its release. It is, however, tied to the Bing search engine which places it more on par with Google (which remains the dominant search engine in today’s market).

In addition to responding to natural speech in several different languages, Cortana carries out a healthy amount of functions including setting reminders, sending texts or emails and playing games.

What Cortana also provide is another player in the increasingly crowded field of digital assistants. No doubt this competition among tech giants will continue pushing the tech forward.

2016 – Bots for Messenger

A year later Facebook introduces a platform providing developers the means to create bots for interacting with Facebook users. Proving popular out the outset, by the end of the year over 34,000 bots were available for application for a wide variety of uses.

2016 – Tay

2016 also brought with it the curious case of Microsoft’s Tay. Its development was highly specific – modeled to imitate the speech patterns and quirks of a teenage girl. At the time, it was an odd yet interesting choice that could have proved capable of providing unique insight into the use of AI.

Tay, however, was a disaster. It morphed into a paranoid and racist tweeter and was shut down less than 24 hours after her launch.

Where Do Chatbots Go From Here

The failure of Tay notwithstanding, the technologies of chatbot AI are rapidly advancing.

As impressive as Alexa may be today, emerging AI like Google’s Duplex offers a glimpse of what’s to come. No longer content to only dial the numbers for a phone call, the Duplex demonstration earlier this year shows AI carrying on an actual conversation.

Thanks to Duplex mirroring the natural mannerisms and tics of typical human speech, the person on the other end remains oblivious they are talking to technology.

It’s not just about a chatbot handling a menial appointment-setting task for us either. There’s already plenty of movement with customer service sectors using chatbots. Soon enough that may expand to more specific interactions such as AI advisors or tutors.

Even after seventy years of advances, chatbots still have the power to wow as they perfect their speech. The closer they get to human emulation, it’s not unreasonable to think of the AI one day performing the Turing test on us.

A Brief History of Chatbots
4.57 (91.43%) 7 votes
Please follow and like us: