Chatbot Intro

In 2017 the conversation is already widely steering to the topic of chatbots. However, for the large part of the public, this is a new notion. So what does chatbot or the chatting robot mean? How does it work? Why do we even talk about it? You are going to get the answers to these questions, in this article.

Bot is the abbreviated version of the word robot. In praxis, however, we use it to refer to robots, which do not materialize. Consequently don’t think about a Japanese waving gadget, rather a virtual robot. They can carry out numerous things, they receive some kind of a signal (input) and the robot, in reaction to this, carries out something (output).

Piecing it together with the word chat, you surely know what is a chatbot 🙂

What is a chatbot?

A chatbot is a computer program which conducts a conversation via auditory or textual methods

This is not so much of a novelty, they had been actively developed since the beginning of the 50s. The most famous first models were ELIZA (1966) and PARRY (1972). These were trying to make the user believe that they are speaking with a real human being. The operating principle of ELIZA is decisive for the developers of the chatbots, up to this day. The bot was able to recognize different keywords and based on those, it generated preprogrammed answers. It was meant to carry on the conversation this way.

However, over many-many years they couldn’t really break into the active every day of our lives. These applications, in many instances, stayed on the level of entertainment, due to their limits.

Fun Fact

My first memorable conversation with a chatbot happened the day before my exam in 2013. Because it was the university’s most difficult subject, I asked it if I will pass my exam the next day

. cleverbot guessing uni exam result

Its guess wasn’t so bad after all: if I had one point less, today I would look back to that one hour, which I spent with chatting, in a completely different way…

Apart from my biggest luck at university, there needs to be a more powerful reason for starting a whole website about this topic. So let’s look into..

Why is chatbot such a hot topic?

trend chatbot 2016

In the last 5 years, the expression „chatbot” was searched internationally like this. At first, I thought that the circled part was caused by Facebook’s announcement at the yearly F8 conference, when they opened the development of Messenger and so they gave a green way for the implementation of the chatbots to Messenger.

But after some research, it seems more likely that the spike was because of the Microsoft chatbot called Tay. It went viral because after 1 day of imitating human behavior on Twitter it became racist. 

I am inclined to think that the Facebook, Google and Amazon trio knows much more about the behavior of the humanity as a whole nowadays, than any other revered expert on Earth. All of the three giants invested a great deal of money into such platforms that help the development of the chatbots and they want to make their own systems, the dominant one.

Why did they choose chatbots?

  • The popularity of the chatting applications surpasses even the usage of the social applications chat applications surpass social appsWe are not just spending a lot of time here but we also communicate, with pleasure, with pictures and with other options which are proposed to us by various programs, apart from writing (for example the like button)
  • The developments in the area of NLP (Natural Language Processing). It is a big challenge for the designers if the user data input is not constrained. In the language, we can express ourselves in many ways, while our intentions behind what we are saying remain the same. For example, for the bot, the two expressions „what hour is it?” and „what time is it?” are not necessarily the same. It has to learn this. This is just a simple example, but you can see that how fast the user can encounter problems if they do not speak in the „language of the machine” and if they do not only choose the opportunities proposed to them by the machine.
  • The processing of the sound-based communication has also developed a lot (mainly in English)
  • The acceleration of communication and the change in our attention span.
    attention span of a human vs goldfish
    The time frame of our attention span is smaller than a goldfish has – we demand a rapid answer and every minute we have to wait for the message, counts.
  • Easy, no commitment starting point for users. Think about the applications, which have to be downloaded and updated. Here it is just two clicks and the update happens immediately.
  • They want you to use their own platforms when you are talking to companies, celebrities or applications. The more time you spend on their platforms, the better it is for them.

So how does a chatbot look like in the real world?

Like how we stated in the definition as well, we can communicate with the chatbots currently in writing, or more precisely visually and with the use of sound.

Sound based communication:

The memorable conversations between Michael and KITT in Knight Rider, which are regarded as humorous by many, today seem very much realistic. Siri, Google Now and Amazon Alexa – just to mention the biggest ones. It helps you to order from the internet, locks the door of the house, turns down the lights or sends the dictated texts on via WhatsApp. This already belongs to the everyday life of more and more people.

kitt as the first chatbot

When is our car going to pick up us when we need it? 🙂

No, your car doesn’t pick you up yet, if you are in a mess, but with the pace of the development of the self-driving cars, even in Hungary (where I’m currently at), this is not so much the distant future. Naturally, we should stay on the ground of reality. Today this is functional, mainly, only in English. If you consider that even KITT from the movie answered sometimes with the expression „Michael this very much confuses me”, maybe the limited conversation, which I conclude with my phone can be better understood.

Visual communication

Here it is easier to process the incoming information for the machine. On the one hand. partly the information is constrained for the available characters. On the other hand, there is an opportunity – based on fix rules – to give possible answers, which the chatbot can understand with a 100% accuracy. The processing of the pictures took giant steps in the past years, but it’s not too realistic to give the robot smarter information with pictures. Due to the fact that in our country Messenger has the most users, I would like to show you an example so you can imagine it better.

One of the ‘oldest’ Messenger bots is the chatbot of CNN, which sends you news on a daily basis. This was presented to me by one of the employees of Facebook, at a conference in Berlin last summer. Since then I have been following the development of the application, and it also observes my taste. Namely based on an algorithm it proposes such articles, which I read with pleasure.

After a preprogrammed prolong it serves the most important news for you according to what it thinks the best suit for you. CNN sends the content in the carousel, but it Messenger chatbots have the opportunity to just send pictures, videos, voice recordings or any other file as well. You can click on buttons with your finger and it opens either its own web browser inside the application, or it sends a summary as a text, directly in the Messenger.

CNN Messenger chatbot

You can try it out by clicking here or just write in the search engine in the messenger the CNN letters. After the start. you can play with it as well.

Now you are familiar with the basics. What a chatbot is, its development and the main directions. Hope you liked the article.
If you are interested in the business use cases of chatbots, register on our upcoming webinar.

Sources:

– Statistic Brain Research Institute 2015, Attention Span Statistics. Available from: http://www.statisticbrain.com/attention-span-statistics/. [14 October 2015].