When You Get An Instant Reply To A Message, It’s Likely The Response Came From A “Chatbots”, Instead Of A Real Person.

When You Get An Instant Reply To A Message, It’s Likely The Response Came From A “Chatbots”, Instead Of A Real Person. Perhaps you’ve dealt with one, without even noticing it. And that’s the whole point: They’re not supposed to be noticed. Just help or serve you; become a “natural” extension of humans. If you deal with a bank, utility, telecom services online there’s a good chance an automated response system is working on your online query.

Welcome to the world of chatbots

Now, there are robots dispensing psychotherapy advice, too, known as virtual conversational assistants (or VCAs), and they’re taking the world by storm. To be sure, chatbots have changed our experience for good. For one simple reason: convenience. Responses come quick and easy. People are generally happy with that. It’s still not perfect, but it works. The result: the likes of WhatsApp and FB Messenger have become key drivers behind chatbots’ rise, and vice versa. Now, more and more are embracing it across the world, which demonstrates AI’s flexibility, cutting through language barriers.

What are chatbots?

Chatbots — chat robots — are personal virtual assistants. They execute instructions in seconds. Consumers like you and me have certain expectations: We don’t want to wait. This is where AI comes in. “One of the most significant AI developments in the last century is the bots and virtual conversations,” Gaurav Singh, founder and CEO of, told Gulf News. “Today, chatbots engage with users in an entirely different way from ELIZA (the first chatbot, which first came out in the 1960s),” he added.

What are they used for?

Chatbots are considered as the “frontliners” of the digital era. They help and are present where human agents cannot be available 24/7. It’s a boon for industries — from banking, law-enforcement, back office, logistics, public services — any entity that deals with people (which is to say, almost all). They are also in travel services, education, utilities, gaming, hotels and even psychotherapy (cognitive behavioural therapy, app which acts like an always-on standby coach suggesting things you must do, from clinically-validated menu of answers).

“Increasingly, they also understand voice commands”, said Singh, whose AI chat platform is based in India. Custom functionality can be built with the AI engine. With superfast computer processing, the system can do things in a flash — answer questions, personalise communication, carry out tasks for the user.

How big are chatbots?

Consider the numbers: One AI-driven chatbot platform (Pandorabot) claims over a quarter million developers and 60 billion messages sent. In 2018, there were 300,000 chatbots on Facebook, according to Venture Beat, business and customers combined sent around 8 billion messages per day in 2020. Amidst the pandemic, chatbot usage increased by 80% (2020), according to Tidio, a chatbot software company. Chatbots are pretty flexible, too: They can support any language.

Can I use a chatbot for free?

Yes. If you’re a start-up, some chatbots allow you to start for free with up to 1,000 messages per month. Some developer platforms charge $0.0025 cents per message with up to 10 bots and 100,000 messages per month.

How does it work?

A chatbot using a natural language processing (NLP) system usually keeps a database of common questions asked that can be referenced, based on key words. So if you have a website and social media presence, a chatbot can boost interactivity: Visitors can get answers to questions about your services or products — immediately — without requiring a human manning a “live chat” to reply.


NLP is language used by humans in everyday conversations. With AI, it also comes with self-learning capabilities that can provide sustained conversational experiences. It does this by gathering huge volumes of customer data, and “learning” from it, in a sort of self-driving vehicle software. Over time, as the AI engine learns using artificial neural networks (ANN) or recurrent neural network (RNN) that can be “trained”, chatbots can increase accuracy of response. This can then increase the effectiveness of organisations. By building a responsive set of answers to frequently-asked questions (this may be done without coding knowledge or the expense of hiring a coder/developer), this innovation is a booster for interactivity.

How did it start?

The so-called “Turing Test” theory was first devised in the 1950s. The first chatbot to be launched was ELIZA, in the 1960s. And if you’ve used Siri or Alexa and their other siblings, you’ve actually become part of the chatbot ecosystem. The Turing Test is a method of inquiry in artificial intelligence (AI) for determining whether or not a computer is capable of thinking like a human being. The test is named after Alan Turing, the founder of the Turing Test and an English computer scientist, cryptanalyst, mathematician and theoretical biologist.


〣 In 2016, both Mashreq Bank and Emirates NBD introduced their chatbots, UAE’s first ever banking chatbots in the Gulf.

〣 Companies like du and etisalat adopted the technology in its earlier years.

〣 The Dubai Roads and Transport Authority uses chatbot to deliver smart Salik services (road-toll system)

WhatsApp’s chatbot is among the top chatbots used for business, according to Sprout Social. Many companies, from travel to event organisers, have implemented this chatbot to improve customer support. Already, more than 80% of small businesses in India and Brazil say WhatsApp helps them improve customer service and grow business.

In the UAE, integration of chatbots with WhatsApp has seen more and more adherents. Travel agents use it to send ticket confirmations as WhatsApp messages — by default. Bookmyshow, an Indian online booking app, does the same. Their users who book tickets are notified via WhatsApp along with the confirmation text or an M-ticket (mobile ticket) QR Code. This allows businesses to revive their approach, boost branding and customer service.

What are the upsides of chatbot use?

  • In general, consumers do like chatbots, according to surveys. Up to 27% of adults — that’s almost 1 out of 3 grown-ups — has used chatbots (specifically for shopping) at least once in their life.
  • Chatbots help improve customer support, save time and resources by automating services and information. And they free human agents to focus on solving more complex problems.
  • Latest data show that 46% of users would still prefer to communicate with a live person. But more and more consumers are buying basic goods, like food or clothes, via chatbots.

Do consumers really like chatbots?

Around 67% of customers used chatbots in 2018, and by the end of 2019 more than a quarter of the population was estimated to be using some kind of chatbot support. Already, around 37% of customers use chatbots to get quick answers in case of an emergency, according to Chatbots Magazine.

How easy or difficult is it to implement a chatbot system?

For first-timers, there are many questions one needs to answer before and during the implementation process. Some of the challenges faced:

  • Figuring out what use cases the chatbot will fulfil
  • Building a bot that brings value to the end-user
  • Designing a bot personality that caters to your audience and matches your brand identity
  • Integrating the chatbot with tech, CRMs and digital channels to provide a seamless and secure experience
  • Knowing at what stage a human needs to be involved.

Q: How do you humanise bots?

Siri (Apple) and Alexa (Amazon) are two examples of chatbots getting “humanised”. Creating characters that represent a brand is not only a PR stunt. It’s based on an understanding of how the mind works: we’re wired to identify with each other.

Brands (or complex computer codes like chatbots) that put a face to a name makes people feel comfortable buying and talking. But there are heaps of ways to use chatbots —  every brand from Pizza Hut to ride-hailing company Lyft and music app Spotify have plunged into the chatbots ecosystem. “By giving your chatbot a character and personality,” explains Singh, “you eliminate the ‘cold connection’ users typically experience where each response feels computerised and robotic.”


〣 A term coined by Uber’s Chris Messina in a 2015 piece published on “Medium”.

〣 It refers to the intersection of messaging apps and shopping. Meaning, the trend toward interacting with businesses through messaging and chat apps like Facebook Messenger, WhatsApp, Talk, and WeChat is more preferred as it encourages buyers to shop online with chatbots.

How do you train an AI chatbot (so awkward or irrelevant answers are avoided)?

The AI chatbot can be “trained” to understand the many ways people can shoot (or even misspell) their questions, explained Singh. “Few ways include defining the use case where each intent contains many utterances — and so on.” Another is by “deep-learning”, for example through the use of Keras, a popular library for neural networks used in building chatbots. One way to start is by creating a yes/no-answering bot. As complexity increases, a programming language like Python may be used to implement a “recurrent neural network” (RNN) structure. “Training” a chatbot can use both supervised and unsupervised machine learning.


Machine learning, a branch of artificial intelligence, is the study of computer algorithms that improve automatically through “experience” and by the use of data. Self-driving technology is one example of machine learning, as continuous application of data helps improves accuracy over time without being programmed to do so.

What languages are supported by chatbots?

Chatbots are pretty flexible, able to support any language. The AI mechanisms behind most chatbots work well with almost every language. Verloop;says they offer support for 18+ languages — English, Hindi, Arabic, Bahasa, Konkani, Tamil, Telugu, Kannada, Hinglish, etc.). They also claim to have a multiple-dialect AI chatbot.

This capability is seen in the spike in demand. reported a three-fold growth in revenue since February 2020. The company has 5,000 brands globally — including Decathlon, Nykaa, Cleartrip, Dar Al Arkan, Fetchr, DSP Mutual Fund, Rentomojo, Scripbox, and others. Amidst the pandemic, chatbot usage increased by 80% (2020), according to Tidio, a chatbot software company. The AI mechanisms behind most chatbots work well with almost every language.

“Our team size has grown three times. The team is now 100+, from 30 members as of February 2020,” said Singh, who expects the company to exit 2021 with five-fold revenue growth. This 2021, the company has new locations launched in Abu Dhabi, Singapore and in the US. Part of its growth plan this year is to hire and relocating tech people to different locations for an Indian global product.


〣TECHVED Consulting

〣 Skcript


〣Dexlock Henote Technologies



〣Q3 Technologies

〣NeuralRays AI Solutions

Is there a FOMO (fear of missing out) syndrome with chatbots?

Singh explains: “It is vital to be familiar with critical new technologies to ensure that you don’t get left behind is also an economic argument. Ultimately, the reasons for using a chatbot are economic, either in the short term or long term. Even the sense of being perceived as a ‘customer’s brand’ for using advanced technology is an attempt to improve the brand — which should lead to more growth.” In many cases, the ROI for chatbots is “exceptionally high”, he said.

What the industries touched by chatbots?

The real estate industry is the most profitable field for chatbots, with more than 28% of real estate business now use chatbots, according to Chatbots Magazine. The travel industry reaps huge benefits from chatbots too. Chatbots, for example, are useful in filtering and speeding up the process of booking rooms or flights. It’s also present in transport/logistics, education, medicine.

In education, chatbots are increasingly used too. For one, they help provide students with personalised feedback that helps improve the overall learning experience. In education, chatbots are increasingly used too. For one, they help provide students with personalised feedback that helps improve the overall learning experience. Moreover, chatbots can recommend relevant online learning content by analysing their learning skills.

Chatbot apps also help teachers in reducing the burden of daily tasks such as checking assignments, evaluating performance, etc. More importantly, chatbots help boost engagement among students, all of whom had been  hemmed into online modes of education. This is how chatbots have increasingly become part of our lives, like the internet itself. At least $5 billion will be invested in chatbots this year, according to Chatbots Magazine.

And thanks to chatbots, the cost reduction of customers across the retail, banking, and healthcare sectors is estimated to amount to $11 billion annually by 2023. More and more people have used channels like Alexa or Google Home chatbot for shopping.

Will chatbots replace mobile apps?

A Gartner report states that many companies are prioritising implementing chatbots over mobile apps. The reason is that. chatbots have become more reliable and effective due to advances in tech and AI. Though this doesn’t mean chatbots will obligerate mobile apps, it simply means when getting quick service, people will look to chatbots first.

This is how chatbots have increasingly become part of our lives, like the internet itself. At least $5 billion will be invested in chatbots this year, according to Chatbots Magazine. And thanks to chatbots, the cost reduction of customers across the retail, banking, and healthcare sectors is estimated to amount to $11 billion annually by 2023. More and more people have used channels like Alexa or Google Home chatbot for shopping.

What does the future hold for chatbots?

In the future, chatbots will be more conversational, i.e. using natural human language. This is the Holy Grail of AI-driven conversation. When close to perfection, with multiple language/accent support, it will enable chatbots to move from simple user-based queries to more advanced predictive analytics based on real-time, human-like conversations, say experts.

Conversational chatbots are already used extensively in many industries and are increasing in popularity. By the end of 2021, almost a quarter of the global population will be using some kind of chatbot support on a daily basis. A Gartner report says that for businesses, AI will be a mainstream customer experience in the next couple of years. It further predicts that 47% of organisations will use chatbots for customer care and 40% will deploy virtual assistants.

So what now?

A Gartner report predicts that 85% of customer interactions will be managed without humans by this year. Is this good news? If you consider the upsides – consumers get answers instantly, companies reduce costs, human agents can spend their time solving more important issues –there’s no doubt chatbots have become one of the most useful tools of our time.

More and more chatbots are appearing. Forbes reported that while millennials use social media apps more than any other generation – they are also chatbot natives. However, it’s the baby boomers (those born between 1946 and 1964) who are likely to expect benefits from chatbots than millennials, say experts. This is because it’s actually the baby boomers who use chatbots to resolve problems more than the younger generations.

Chatbots have an inherent utility. No matter whether or not we prefer talking to agents, in the future people will be forced to rely on chatbot support, just like we rely on apps now. Chatbots will change the world, one chat at a time. This is almost a certainty.


1950: The Turing Test

Alan Turing theorised that virtually intelligent machine would be indistinguishable from human during a text-only conversation. This idea essentially laid the foundations for the chatbot revolution.

1966: Eliza

Eliza was created at the Massachusetts Institute of Technology (MIT) AI lab to simulate human conversation by matching questions with the scripted responses. It gave an illusion of understanding but had not been in framework for contextualising events.

1972: Parry

While Eliza was a tongue-in-cheek simulation of a therapist, Parry simulated a person with paranoid schizophrenia. It used a conversational strategy and was much more serious than Eliza.

1984: Racter

Racter is an artificial intelligence computer application that generates English language prose at random. It was published in 1984 by Mindscape.

1988: Jabberwacky

Developed in the 1980s and released online in 1997, Jabberwacky chatbot was designed to simulate natural human chat in an “interesting and entertaining and humorous manner.”

1992: Dr Sbaitso

An AI speech synthesis program created for MS DOS based PCs. Dr Sbaitso assumed the role of a psychologist when interacting with others and was designed to showcase a digitised voice.

1995: Alice

The artificial linguistic Internet computer entity was a “natural language processing bot”. She could apparently follow matching rules based on input in order to have conversation, but was still unable to pass the Turing test.

2001: Smarterchild

An intelligent bot widely distributed across SMS networks and friend lists among AOL and MSN messenger users. It offered a personalised conversation and was considered a precursor to Apple’s Siri and Samsung’s Voice.

2006: Watson

Watson was originally designed to compete on the TV show “Jeopardy” in which he beat two of the show’s former champions. Watson has since gone on to bigger and better things using natural language processing (NLP) in machine learning to reveal insights from large amount of data.

2010: Siri

Siri is an intelligent personal assistant. It is part of Apple’s IOS and uses a natural language user interface to answer questions and perform various requests. Siri did the groundwork for all later AI bots and personal assistants.

2012: Google Now

Developed by Google for a Google search mobile app, it employs a natural language user interface to answer questions, make recommendations, and perform actions by passing on requests to a set of work Web services.

2014: Slack Bot

Slackbot is a built-in personal assistant on Slack (a channel-based messaging platform launched in 2013). Slackbots are handy assistants: they wait for commands, then find or create the thing you need. You can also add some character to your conversations by producing custom responses for Slackbot.

2015: Alexa, Cortana

Alexa is an intelligent personal assistant that inhabits the Amazon Amazon ecosystem. It is capable of voice interaction, using natural language processing algorithms to receive, recognise, and respond to voice commands. Cortana is Microsoft’s version of the Intelligent Assistant that can set reminders and answer questions using the Bing search engine. It recognises natural voice commands and in a number of languages.

2016: Bots for Messenger, Tay

In April 2016, Facebook launched a messenger platform is which allows developers to create bots that can interact with Facebook users. At the end of that year, 34,000 bots were available covering a wide range of use cases. Tay was the chat bot created by Microsoft to mimic the speech and habits of a teenage girl. It caused controversy when it began to post offensive tweets and became increasingly paranoid. It eventually had to be shut down 16 hours after launch.

2017: Woebot

It is an AI-powered Cognitive Behavioral Therapy (CBT) chatbot that delivers a suite of clinically-validated therapy programs to address mental health challenges. Woebot functions as a coach who chats with you and offers insights and skills to help you grow into your best self. You can chat with Woebot as much or as little as you like — it’s always available when you need it.

This news was originally published at Gulf News.