Over the past few months, artificial intelligence (AI) appears to have gone from being the preserve of researchers and major technology firms to something that’s accessible to everyone with an internet connection. Unsurprisingly, a growing number of businesses are looking for ways to integrate AI tools into their business systems and offerings.

A Gartner survey released earlier this year found that 79% of corporate strategists view AI as critical to their success in the next two years. But amidst all the hype, how helpful is AI really? And what parts of an organisation is it best incorporated into?

According to Euphoria Telecom chief technology officer Nic Laschinger, there is no doubt that AI-enabled tools can be incredibly powerful in a variety of business applications. But, he cautions, that’s only true if they’re used correctly.

“Small businesses especially often lack the resources to bring in the expertise they need to have full-scale marketing, sales, and research departments,” he says. “Tools such as Simplified, Quillbot, and Grammarly can help ensure their marketing copy is clean and compelling while Hubspot and Salesforce have AI analytics and machine learning baked in to give businesses a better understanding of customer behaviour and sales trends. Large language model (LLM) tools like ChatGPT and Google’s Bard, meanwhile, are potentially transformative for research.

“As helpful as AI tools can be, however, they need to be used carefully,” he adds. “Some AI tools occasionally have ‘hallucinations’, meaning that they generate false answers.”

Those hallucinations can be costly too. Security researchers, for example, have found that ChatGPT hallucinations can be exploited to distribute malicious code to software developers. That can then lead to data breaches, which now cost an average of R50-million.

“That’s just one of the reasons why companies shouldn’t use AI to replace any positions,” says Laschinger. “Instead, they should look at how AI tools can complement their existing resources and make their jobs easier.”

That’s the case for more specific business applications such as telephony too. AI can, for instance, help automate tasks such as call routing, appointment scheduling, and basic customer queries. That, in turn, reduces the workload on human operators and improves efficiency.

As Laschinger points out, AI can also improve the overall customer experience.

“AI tools, implemented correctly, can feed telephony providers personalised recommendations based on the individual data of a particular customer,” he says. “This personal touch, along with the ability to more quickly resolve customer queries, can enhance how customers experience the company, building loyalty and trust.”

If a business has a contact centre, AI tools can analyse call data and provide further insights into things like customer preferences, behaviour, and sentiment. These can then be used to further improve the customer experience.

“On the other side of the coin, if a business tries to use AI to replace contact centre operators entirely, it can cause rapid damage to the customer experience,” says Laschinger. “Think about how many times you’ve tried to resolve an issue with a company only to tear your hair out in frustration because you can’t get any human help.”

It’s clear that while AI can be incredibly helpful to businesses and is already being effectively deployed by many companies, it should primarily be seen as a tool rather than an all-encompassing solution.

“The recent advances in AI have been nothing short of extraordinary,” says Laschinger. “But for all those advances, AI still works best in tandem with skilled human workers rather than as a replacement for them.”