What should it feel like to have your apps altogether just in one single intelligent conversational interface? Literally just ONE app for everything. Can you imagine that? Let’s trace together with the way from a simple device (where we all started) to smart virtual assistant.

Since the 2000s when our cell phones started their transformation into something way more complicated than the just device for calling and sending messages, we have been crazy about constantly evolving functionality making our life easier and more pleasurable. Now when we’re going to get a new model of favorite smartphone, it has nothing to do with its innate functions – we’re buying digital factories which are dedicated to helping us with whatever we’re doing. But let’s think a little bit more about “app revolution”.


After first successes app market began to grow in never seen before (in the digital world) rapid way. It seemed like everyone started to talk about new application. It was clear that demand was enough for everyone who was able to produce a sophisticated application to fit in this market.

Developers expected that for every single need from waking up through ordering food through buying things to planning a vacation should be its app, so to rule our life we would need to push the buttons and that’s it. However, in a while, it turned out that even though new apps still used to draw attention, users were not as active as predicted.

Following surveys revealed that average person has in active use up to three apps through the day, among which the most popular were messengers. From small to giant companies realized that it’s actually not needed to produce more apps but to come up with the single one possessing easy for user conversational interface (e.g. messaging, talk).

First of all, we’re humans and we are actively interacting with the environment by means of natural language. Even the fact that our technology can propose us much more functionality than we naturally have, most of the time we want to use natural way of interaction. That’s how first chatbots appeared.


Now I propose you to deep a bit into the terminology of robots, bots, chatbots and conversational interfaces. Let’s figure out main differences between them to move forward with it.

The term “robot” was invented by Joseph Čapek and firstly appeared in his brother’s (Karel) play in 1920. However, first trials of crafting automated machines date back to 400 BC. Definition of the term combines electromechanical devices resembling some extent human appearance and virtual software agents operating over the Internet, each of which aims at performing repetitive automated actions. Becoming more and more popular the latter (virtual software agents) have lost first few letters and now are widely called “bots”.

What about chatbots?

The main difference of chatbot lies in its conversational component which makes them so demanded. It’s more than just one more improvement like those permanently appearing in new applications – but it’s an answer to our natural need for natural communication. The most widespread examples you might have already come across are online store virtual assistants.

Bots also live on Twitter and write back to you when you follow somebody. They’re working on Facebook business pages and give you information on FAQ when you come up with some question to ask a business representative. The bunch of them inhibit Skype and at any moment are ready to consult you on a variety of questions – from weather to psychological problems.

Let’s look at the most interesting recent examples of chatbots, though with somewhat different outcomes.

1) XiaoIce

Virtual- china- xiaoice-chatbots-evolution

XiaoIce is chatbot hit among Chinese customers developed by Microsoft Corporation. It’s more than virtual assistant like Cortana – but it’s emotional companion as most of the users themselves describe it. This’s how Microsoft commented on their new invention:

By simply adding her to a chat, people can have extended conversations with her. But XiaoIce is much more evolved than the chatbots you might remember. XiaoIce is a sophisticated conversationalist with a distinct personality.  She can chime into a conversation with context-specific facts about things like celebrities, sports, or finance but she also has empathy and a sense of humor.

Using sentiment analysis, she can adapt her phrasing and responses based on positive or negative cues from her human counterparts.  She can tell jokes, recite poetry, share ghost stories, relay song lyrics, pronounce winning lottery numbers and much more. Like a friend, she can carry on extended conversations that can reach hundreds of exchanges in length.  

Now XiaoIce is being used in China by more than 40 million people.
24-hours available online friend always willing to hear from you, whom you would want to tell sometimes more than a real person. Sounds pretty good, right?
However, not all of our trials to make the smart bot which acts like a human are so positive.

2) Tay

Tay was an artificial intelligence chatterbot released by Microsoft Corporation on March 23, 2016. Tay caused controversy on Twitter by releasing inflammatory tweets and it was taken offline around 16 hours after its launch.[1] Tay was accidentally reactivated on March 30, 2016, and then quickly taken offline again.

Tay was AI-driven Twitter chatbot released on 23 of March 2016 by Microsoft Corporation. She was invented to mimic 19 years old American girl conversational behaviour targeted at engagement and entertainment of 18-24 years-olds.  Just after 16 hours online under influence of trolling attacks she began to release racist&sexist inflammatory tweets and was turned offline. On 30 of March 2016, Tay was reactivated.

What went wrong with Tay?

Don’t forget that Tay possessed NLP (Natural Language Processing) abilities. And since she was able to reply to Twitter messages and caption photos tweeted at her, she was literally learning from all this information addressed to her. Being attacked by trolls messages with a variety of inappropriate content, Tay learned from it and just repeated some of them.

Why do we need to know about such polar examples?

Regarding alternative of achieving artificial intelligence when a machine will be able to learn thousand times more rapidly from given data that any human, we need to develop strong security rules. Otherwise, it may result in dangerous circumstances.

Here is what Microsoft Corporation wrote in their blog after:

“Looking ahead, we face some difficult – and yet exciting – research challenges in AI design. AI systems feed off of both positive and negative interactions with people. In that sense, the challenges are just as much social as they are technical. We will do everything possible to limit technical exploits but also know we cannot fully predict all possible human interactive misuses without learning from mistakes.

To do AI right, one needs to iterate with many people and often in public forums. We must enter each one with great caution and ultimately learn and improve, step by step, and do this without offending people in the process. We will remain steadfast in our efforts to learn from this and other experiences as we work toward contributing to the Internet that represents the best, not the worst, of humanity.”


Regarding open development platforms and a variety of easily accessible tools provided by such companies like Facebook and Microsoft today is as never before we’re encouraged to make users comfortable with creating a highly personalized experience. That’s what conversational interface is in general. Looking closer we can find out that primarily it’s about easy and natural interaction.

What’s next?

Intelligent Conversational Interface is what we are working on. You can say: “Well, I thought we already had CI”. It’s true we already have virtual agents around us. We can easily contact them at any time by means of natural language and they will understand our inquiries nearly as good as a human would.

Yet, the main problem is that they’re good in handling tasks in the strongly defined field. In other words, they have narrow specialization. In its measures, they can learn and improve but they won’t step outside. ICI in contrary is about multitasking and it’s what the biggest scientific minds are working on.

To give you more insight about the principal difference of CI and ICI I decide to come up with some examples which some of you might already know.

Mr. Robot

It’s the scene of American TV series “Mr. Robot” where Dom (Dominique DiPierro) is talking to her virtual assistant Alexa. Alexa can wake Dom up, can answer on something like “When is the end of the world”, can recommend what to wear due to the weather forecast. Alexa is pre-programmed to answer questions which are out of range of its competence (“Do you have a boyfriend?” or “Do you love me?”) in the very same way for every user.

Is it helpful? For sure, to some extent. Is it real? Yes. It’s Amazon’s productand it’s roughly what CI is now.



Samantha is intelligent software operating system in American drama “Her” (2013). After the protagonist of the movie, Theodore, answered set of questions on his personality, he received this system adapted personally to him. Samantha actively learned from their mutual conversations, from Theodore’s preferences, habits, behaviour. She had an appropriate sense of humour, she could advise whatever question Theodore could come up with, therefore, she became better and better fit to exactly Theodore’s personality.

Is it helpful? Might be extremely. Is it real? Unfortunately, not yet. This is just one possible way of how ICI can look/feel like.

As we might have already gotten more or less clear understanding of the point where we are staying now,  it’d be useful to take a quick look at the road we’ve made so far.

The chronology of chatbots evolution


1950 – traditionally tracing the history of AI studies starts from “Imitation Game” proposed by Alan Turing, which is actually the test for a machine which includes a set of questions passed to the machine and to human by (human) interrogator without knowing exactly who’s currently answering. In case if interrogator concludes that machine is of human origin, it would mean that the machine successfully passed the test and it possesses intelligence.


1966 – first chatbot ELIZA was invented by Joseph Weizenbaum to emulate psychotherapist’s conversational patterns when talking to the patients.

ELIZA and PARRY met twice.

PARRY was an early example of a chatterbot, implemented in 1972 by psychiatrist Kenneth Colby


1972 – chatbot PARRY was written by psychiatrist Kenneth Colby and was to simulate patient with paranoid schizophrenia.


1981 – chatbot program Jabberwacky was created by Rollo Carpenter and had a primary focus to pass Turing test. It was written to mimic human interactions for entertainment purposes only. You can try it here.


1992 – artificial intelligence speech synthesis program, called Dr. Sbaitso was released by Creative Labs. The idea was that the program had to act as a psychologist. Underlying algorithm resembles the one ELIZA is based on.


1995 – the winner of  Loebner Prize A.L.I.C.E. (Artificial Linguistic Internet Computer Entity) was originally written by Richard Wallace. Receiving message as input it applies heuristic pattern matching to it. In 1998 A.L.I.C.E. code was rewritten in Java and now is available as open source project.
Siri is a built-in "intelligent assistant" that enables users of Apple iPhone 4S and later and newer iPad and iPod Touch devices to speak natural language voice commands in order to operate the mobile device and its apps.


2001 – a precursor of Apple’s Siri and Samsung’s S-Voice SmarterChild was invented by ActiveBuddy. It was available AOL Instant Messenger and Windows Live Messenger and gained a lot of users truly attached to it (around 30 million). It was acquired by Microsoft in 2007.


Watson is an artificially intelligent computer system capable of answering questions posed in natural language, developed in IBM's DeepQA project by a research team led by principal investigator David Ferrucci. Watson was named after IBM's first CEO and industrialist Thomas J. Watson.
2006 – IBM’s Watson was initially invented to win game Jeopardy! and in 2011 it accomplished this goal. Now it’s being used to process huge amounts of data and make predictions.

2010 – Siri is an intelligent virtual assistant and a part of Apple iOS. It’s working by delegating user’s inquiries to Web Services.


2012 – Google Now is an intelligent virtual assistant available on Android and iOS which also uses Web Services.


2014 – Chatbot Eugene Goostman crafted by two Russians and one Ukrainian has passed the Turing test for the first time.


Alexa, the voice service that powers Echo, provides capabilities, or skills, that enable customers to interact with devices in a more intuitive way using voice. Examples of these skills include the ability to play music, answer general questions, set an alarm or timer and more. Alexa is built in the cloud, so it is always getting smarter. The more customers use Alexa, the more she adapts to speech patterns, vocabulary, and personal preferences.

2015 – Amazon Echo device’s virtual agent Alexa; released by Microsoft intelligent personal assistant Cortana, which recognizes voice commands and uses Bing search machine to provide users with requested information.


The Messenger Platform gives you the ability to have conversations with people on Messenger. We've added new tools for you to build and promote your bot so you can create a custom experience for your unique audience.
2016 – Finally it’s time for bots to step out from the huge companies and gifted minds. Microsoft and Facebook created open platforms that are full of appealingly easy tools for everyone who is willing to contribute to the era of automation.

The more complicated technology becomes the wider range of tools we can access in order to create the unique user experience. The story of chatbots evolution naturally comes to the point where we now are, when we don’t perceive it as an outstanding tool rather as the one that saves you more time to actually live, the one that we’ll be gradually improving to get more from it. This kind of approach is what we’re evolving toward whole AI branch. It’s exciting times we live at. Just as much exciting content is going to be here.

Recommended reading:

Top industries to use chatbots

Are the chatbots important for current technology era? Chatbots are obviously the main tech trend we’re talking about in 2016.

How to build smart customer service

Build smart customer service. Most of us imagine that in the future we all will live and work in smart houses, surrounded by robots.

Pros and Cons of Chatbots

Education, marketing, healthcare, lifestyle, customer services – all of these fields suddenly turned out to be divided into automated tasks with chatbots.

Growth Hacker and Sales Hacker, MVP builder, love to run technology companies.