Chatbot vs conversational AI: What’s the difference?

THERE'S a good chance that by this time, you've used or interacted with an AI, or a “chatbot”. You may have unknowingly dealt with a number of them. Chatbots are everywhere, and have become part of our lives.

Disclosure: A chatbot was engaged in prepping for this article too (Say: “Hey Siri/Alexa, what is a chatbot?”). In reality, chatbots are virtual beings. Software "trained" to respond to humans. And they work quite decently, 24/7 without complaints or asking for health benefits. Therein lies its beauty — as well as a very tough challenge.

However, engaging in human-like conversations is a tough one, even for the smartest AI engineers. That’s why there an emerging dichotomy between conversational AI and chatbots.

First, let’s deal with the chatbot part. A chatbot is more than just kiddie fun. It is an industrial, practical use of artificial intelligence (AI). It chats with you, gives you information, supports, shops and books things for you.

But it can do so much more. In general, chatbots are modern-day digital assistants and have proven pretty useful. They are great at answering routine questions or do routine tasks perfectlly without complaints or off days.

Thanks to AI and “machine learning”, chatbots are “learning”, becoming more conversational. Industry experts say this will raise the use of chatbots to a higher pitch, though it’s still a long way before conversational AI can achieve human-like problem-solving capabilities.

Chatbots are popular, and that’s increasing by the day. They come in different shapes and names. Siri, known as the first intelligent assistant for everyone, came out in 2011; Alexa in 2014. They’re just two examples.

Siri brought AI to the mainstream. It’s now an integral part of Apple products. It is used more than 2 billion times a week in over 30 countries around the world.

Alexa, by Amazon, is another example. Amazon data show there were 53.6 million Amazon “Echo" speakers sold in the 2020. The number grew to 65 million in 2021.

Unfortunately, chatbots or digital assistants can be unpopular with many customers.

Chatbots can be frustrating, too, or serve downright foul goals, such as transmitting computer viruses or artificially increasing views on YouTube videos or web articles.

Also, the use of text on a screen is quite limiting. Messaging or voice-based applications are the new frontier, taking chatbots to the realm of ubiquitous, unobtrusive — and “human-like”.

WhatsApp, one of the most-used apps, is among the top chatbot platforms used specifically for business today. Software firm Sprout Social estimates that 80% of small business owners in India and Brazil said WhatsApp helps them improve customer service and grow business.

More and more bank interact with customers on WhatsApp — giving account statements and the like, for example. It’s convenience is thanks to the underlying AI.

Now get this: Today, there are tens of thousands of these digital assistants in use. That’s only bound to increase. Why? They’re quite useful. Back in 2018, there were already more than 300,000 Facebook chatbots.

That same year, the number of chatbots increased by 100,000 users. Business and customers combined sent around 8 billion messages a day over the past year. These staggering figures are only bound to increase.

What’s the way forward for chatbots? Usage is only expected to spike as the technology matures.

A Forbes report shows 3 out of 5 millennials — this digital-native generation known for trying new technologies and applying them everyday — have used chatbots at least once in their lives.

Millennials (or “Gen Y”, people born between 1981 and 1996) use social media apps more than any other generation. They interact with friends, get information, shop, or use apps for entertainment.

It’s not just the younger crowd: When “baby boomers” (the generation born between the end of WWII and the mid-1960s) use chatbots, they expect them to resolve problems, with around 61% expecting an “instant response”.

This is a high bar. And it’s a complicated by the fact that people also expect more “humanised” chatbots. It’s a tough challenge, but also presents a huge opportunity for the chatbot service community.

What's the biggest challenge for the chatbots industry. How difficult it is to train the underling AI? For Werner Kunz-Cho, CEO of New York-based Fareportal, the tech firm leading hybrid travel agencies like CheapOair, OneTravel, the biggest challenge for chatbots in their industry is to distinguish between “intents”, given similar words and phrases.

“For example, a customer typing, 'I want a refund’ can be translated to either customers wanting to cancel their reservation or customers wanting an update on a refund for a reservation that was cancelled previously. This, coupled with understanding the sentiment of the customer, further makes it more challenging.”

But for every problem there is a solution, or at least a roadmap for one. Kunz-Cho cites 3 key ingredients for chatbots can really help boost customer experience: conversational flow, 24/7 availability and deep integration.

“Cognitive chatbots” allow for greater business integration and presents a way forward. In the airline industry, for example, cognitive chatbots deliver end-to-end airline reservation process.

For simple airline bookings, the chatbots can and do perform this function.

But there are still a lot of complex scenarios where a chatbot may not be able to guide a customer properly. So, there is still a lot of scope of improvements there and everyone is working towards plugging that gap.

“Another challenge in getting this done is the human understanding of technology. As a user, if I understand how these ‘cognitive chatbots’ work and respond, I can get the best out of it. But ‘solutioning’ it for a non-tech savvy user, it is going to be a challenge,” Kunz-Cho said.

Quick answer: Yes. Shopping bots have been a boon to retail. More people expect being able to buy basic goods, like food or clothes — i.e. coconut flour, nuts, even luxury goods — online.

Chatbots Magazine estimates nearly 13% of shoppers are now buying more-expensive items using chatbots.

Tidio, a chatbot provider, reported that women tend to use chatbots for online shopping. On the other hand, men are most likely to turn to this technology to get answers to simple questions: 47.92% of men use chatbots if they can’t find answers to simple questions.

One industry estimate projects a 22.5% annual growth for chatbots, from $400 million in 2019 to $1.95 billion by 2027.

It’s a system where chatbot and human agents work together to improve the customer experience.

The bot will be able to handle more and more chats. When they transfer the chat, they provide complete context to the agents for resolving the customer query quicker and better.

Human touch would always be essential part of the customer experience, argued Kunz-Cho.

While AI-powered chatbots engage your customers in real-time —by making live chat widgets easily-accessible to customers 24/7.

“A customer who is frustrated or stuck in a complex problem would need to talk a live agent. But the overall efficiency of the support system will keep on increasing with automation being made available through interactive chatbots.”

Could an article like this can be put together by a chatbot, and in a flash and make it more engaging? Perhaps. It could already be happening somewhere. And less boring, too.