Conversational artificial intelligence (AI) space is brimming with a lot of action these days, thanks to the popularity of Microsoft-backed ChatGPT. OpenAI’s ChatGPT has kicked off an innovation race that will accelerate technological developments much faster than before, says Adit Jain, co-founder and CEO of Leena AI, which announced the release of an integration with GPT-3 (the backbone of ChatGPT) with its AI tool to enhance end-user experience. In simple terms, this integration will help reduce the number of unanswered employee queries, improving the overall end-user experience of interacting with the Leena AI dashboard.
“AI chatbots have proven to be a game-changing tool for improving workplace collaboration and creating a happy culture,” says Vikas Kakkar, founder, Amara.ai, an interactive HR tech startup with an AI chatbot. “For example, Amara.ai acts as a virtual buddy for employees, and when it is integrated into people’s practices, the entire engagement process becomes a seamless and hassle-free experience.”
It is worthwhile to note here that although the OpenAI chatbot’s answering abilities shocked many, conversational AI has been around for a while now and is already transforming markets across different industries. Where has this niche technology reached so far in terms of adoption, what are the major roadblocks here? Most importantly, does it lead to the loss of human jobs?
The present scenario
A recent report by Gartner estimates the conversational AI platform market at around $3.8 billion, growing 55% year-over-year. At the moment, chatbots are the most widely used kind of AI in businesses. Over the next two to five years, their adoption rates are predicted to nearly double. By 2023, AI-powered chatbots are expected to save upto 5 billion work hours. Contact centre deployments using conversational AI will save labour expenses by $80 billion by 2026.
“Conversational AI is, however, not immune to challenges. It must adapt to new speech trends as human language changes,” says Rashi Gupta, co-founder & chief data scientist at Rezo.ai, a Noida-based firm that uses machine learning and natural language processing (NLP) to automate enterprise workflows with limited human intervention. More specifically, it is an AI-powered platform that automates conversations over voice, emails, WhatsApp, social media, and chats.
In her opinion, conversational AI is evolving as one of the substantial applications of AI and is the most manifested kind in mainstream/ day-to-day usage of people’s life. With time, one can see lines blurring between conversational AI, communications platform as a service (CPaaS), and contact centres verticals. The key performance indicators and cores of contact centre verticals that we see today were all designed over the last 10-25 years beginning with what was available at the time.
Over the course of time, layers of tools/ features/ capabilities have been added to inmprove measurability, efficiency or performance. If these indicators were to be designed today, given the technology, cloud, and cost structures that we have, they shall look very different.
Conversational AI, if leveraged in the correct way, is that window to possibilities — the potential which becomes much larger when seen in tandem with data analytics, a very essential piece. “Once we integrate data analytics with automation, we see a range of improvements in terms of learning, improving, and unlocking new growth levers to maximise revenue for enterprises other than augmenting the painstaking agent tasks. Conversational AI has transcended its applicability and is cutting across not just different industries but multiple departments within the organisation,” says Rezo’s Gupta.
Closer home
India is home to 22 officially recognised languages and an astronomical 19,500 dialects. Hence, building and implementing a conversational AI solution that is capable of speaking to customers and registering their responses in their own dialect will remain a challenge. “However, we, at Rezo, have always taken this challenge head-on. After being there in the market for more than 4 years, we have built sophisticated language models trained on more than 20 million data points which are able to understand multiple dialects across 10-plus languages,” she says.
Also read: Hey Elon Musk, did you just call yourself truly corrupt on Twitter?
Market acceptance
Conversational AI is getting accepted across industries, especially in B2C verticals where it is offering better efficacies in larger volumes. For example, by inculcating AI and automating contact centres, banking and financial services can be made a readily available commodity for the masses. In the telecom sector, conversational AI can have a positive impact by enabling mass reach-outs and preventing funnel leakage by providing better offers and running dormant reactivation campaigns.
“We have seen that even smaller companies, with more experience in handling consumers, were able to extract the maximum benefits of conversational AI while large enterprises with lesser experience in consumer handling struggled with making it work. However, a host of other verticals, other than B2C, too are adopting conversational AI at a rapid pace. The average deal size for conversational AI solutions increased by 28% in the last year alone, which signifies a phenomenal scale-up and growth into multiple verticals resulting in its much-expanded usage,” points out Gupta.
Threat to human jobs?
Also read: Oppo Find N2 Flip launching globally on February 15; Samsung Flip 4 rival might just come to India
Historically, new or better technology has never led to job losses. On the contrary, it has expanded opportunities for all. There is a difference between job loss and the need to up-skill – which often gets mixed up in this context, says Gupta. “At Rezo, we don’t see technology as a replacement for human resources but rather as a solution to augment human productivity and as complementary to each other.”
The bottom line: Conversational AI is enabling human agents to do what they do in a much more efficient manner. And being a subject of many a chat at the same time.
