AI Trends: Can ChatGPT Revolutionise the Contact Center?

Artificial intelligence has the potential to fundamentally change the contact center. Shorter response times for customers and lower operating costs are just two of the many benefits it can bring. However, the reality is that only a small percentage of contact centers worldwide are benefiting from these advantages today. But why would companies delay an AI integration when the added benefits are so great? The reason is that training and maintenance of AI models is expensive and must be customized for each contact center. Although these costs have dropped significantly over the past decade, they remain prohibitively expensive for many.

In the past two years, there have been major developments in AI models for human natural language processing as required by contact centers. We are now in what is called the Large Language Model (LLM) phase. The best known of these models is GPT-3 from OpenAI, but there are other competing products and open source variants. These models eliminate the need to learn the AI and allow the functionality of the model to be customized. This, in turn, means that the cost of training and maintaining AI models drops dramatically. Deploying this technology in the contact center industry removes the last barrier to mass market AI adoption.

ChatGPT can revolutionise the contact center

Chatbots are, at first glance, an obvious application for ChatGPT. However, they are not the first-choice application areas because ChatGPT can answer questions today, but it is not capable of performing actions. When a customer contacts customer service, they are looking for support beyond a simple answer. In many cases, these are more complex tasks such as processing a return, canceling an account, or transferring money. Will this eventually be possible? Undoubtedly.

When ChatGPT is used to answer questions, it answers based on information available on the Internet. However, it does not have access to offline knowledge. Often, customer queries involve information that cannot be found on the Internet. This is the very reason why customers often call customer service in the first place. ChatGPT excels at generating text and creates new content from existing online information. When a user contacts a brand, they don’t want creative text, they want immediate action. All of this will be possible in the future, but it also means that the first use case will likely not be chatbots.

Sensitive customer trust

If companies try to pass off automated voices as human contacts, it leads to a loss of trust. No customer wants to feel deceived when they realize they’ve been communicating with an AI. In 2023, the customer experience will determine the success or failure of brands. People are turning to customer service more than ever with emotional and stressful issues such as rising costs of living, travel disruptions, and extreme weather conditions. In the face of such issues, which could intensify this year, those brands that demonstrate empathy and understanding – aspects that often go hand-in-hand with interpersonal contact – will win the race.

In 2023, therefore, companies will work toward a better balance between human and digital customer experience. Human empathy and judgment combined with the speed and data breadth of an AI are critical to managing the increasing volume of sensitive customer contact. This shift toward collaborative and intelligent human-AI collaboration will help empower employees while enabling greater self-service and customer satisfaction.

Build instead of buy

Until now, contact center software vendors developed their AI solutions entirely in-house. They collected data from customers and used it to train, customize, test, optimize, and maintain AI models. The conventional wisdom was that they would only survive in the AI space if in-house development came first and acquisitions came second. Many of the major technology vendors, such as Amazon, Google and IBM, are currently doing important work. They can recoup high AI development costs by selling into a variety of markets beyond the contact center market. Companies are no longer forced to spend long hours developing their own solutions, and this mindset will continue to spread in times of recession as executives seek to cut costs.

In 2023, the trend will intensify accordingly as technology becomes more complex and powerful. Teams will focus on innovating using the technology available, rather than spending valuable time and resources developing custom AI models for customers and collecting big data. This will set a new benchmark for the contact center industry, giving it a significant improvement in its working foundation.

Jonathan Rosenberg

CTO and Head of AI, Five9