Introducing the New AI on the Block: Generative AI 

What’s all the hype around artificial intelligence? Hasn’t AI been around for years? 

No, you’re not crazy. 

AI has been around. For years, it’s been auto-suggesting Google search queries and giving you suggestions for an email response. 

But that’s the old AI, which relies heavily on predefined rules to do things like make simple decisions and recommendations from your existing help center content. They’re limited to the knowledge explicitly programmed into them and lack the ability to learn or adapt from data. Traditional machine learning techniques could make predictions or classifications based on data patterns. Techniques like Natural Language Processing (NLP) could parse text, conduct sentiment analysis, and retrieve information. 

But none were capable of generating human-like text or content on its own. 

If the old AI felt like a GPS that got you from your home to a new restaurant, then the new AI feels like a self-driving car that knows every short cut, pothole, and can have a conversation with you about the weather. 

OK, that might be an exaggeration. But tech forward-thinkers like Bill Gates have even said the new AI is as revolutionary as the Internet and mobile. 

The new AI is Generative AI, which can independently create new content, realistic images, human-like text, and new music. These new AI models learn from large datasets, then replicate the patterns and structures from the training data to produce new content.

The brains behind Generative AI are Large Language Models (LLMs). These models have been trained on massive amounts of text data (trillions of words, sentences, and concepts) from books, articles, websites, and other sources. 

They can learn language patterns, grammar, semantics, and knowledge about the world from this data. They leverage their general language understanding abilities to perform specialized tasks with relatively small amounts of task-specific data (compared to the old AI). 

One of the most well-known LLM is OpenAI’s GPT 3.5 which powers ChatGPT, a text chatbot released in November 2022 that can have natural conversations with humans. Other leading foundational model companies, such as Google, have introduced LLMs like LaMDA and PaLM, while Meta has launched LLaMA. These models are intricate and costly to operate. Prominent companies in the LLM space include OpenAI, Cohere, Anthropic, Google, Meta, Amazon, and NVIDIA, with a select few offering their LLMs for broader commercial use.

Applications like DALL-E, the AI image generator, harness the power of a Multimodal Model (which encompasses both language and imagery) via APIs to provide useful tools in a friendly interface. While DALL-E does utilize aspects of GPT-3 by OpenAI, it's not solely powered by an LLM. Today, there are hundreds of AI tools available to assist with writing, generating ideas, searching, developing code, extracting data, and many other tasks.

The potential uses of this Generative AI technology are massive. From consumer to business use cases, the AI software market is growing fast – analysts predict the market will grow to $1.3 trillion over the next 10 years. It’s clear that we’re just at the starting line of this AI revolution, and as a customer experience leader, you can’t afford to be in the dark.

A new chapter for customer experience with generative AI 

Historically, AI has had a bad reputation in customer service. 

A chatbot that struggled to process a basic refund. A voice system with so many prompts that it drove customers crazy. 

The old AI technology also required months of heavy upfront work and ongoing investments in setting up the rules, content, and information systems. And for what? Lots of empty promises and unmet expectations.

Generative AI, on the other hand, is showing a lot of early promise, especially in driving cost-efficiency for customer experience teams when used properly. It’s capable of:

  • Improving average handle times and first call resolutions by better assisting agents 
  • Deflecting customer contacts with more reliable self-service support 
  • Producing content in multiple languages at extremely low costs
  • Uncovering insights on how to improve your workflows, agents, content and more 

However, customer experience leaders widely agree that generative AI solutions work best alongside humans, and not in isolation as a complete replacement. 

For example, the ability to generate original content puts companies at risk without proper oversight on what AI may say to a customer. More importantly, AI lacks the human touch. While it can simulate empathy or understanding, it can’t actually feel anything. Leveraging Generative AI in customer interactions without human oversight creates the risk for misunderstanding and disconnection.

With generative AI, the potential to help businesses scale and improve their customer experiences is real. But it requires the right leader and the right mindset to implement it successfully. 

“A lot of AI systems such as chatbots and IVR systems are bought for cost-savings, not necessarily user experience. We have to try not to lose that North Star. What we’re actually trying to do is help the customer.” - Mariliam Semidey, Director of Support & Localization, PicsArt Inc.

The real uses of generative AI in CX in the wild

ChatGPT made access to AI solutions so ubiquitous that families are even using it to plan dinners and grocery lists. 

Although that’s true, we’re still in the early days of using generative AI for business and customer experience use cases. These solutions are growing by the week as AI technology gets integrated into existing contact center platforms and as new AI startups get funded. 

Here are some of the ways Generative AI is improving the customer experience and driving cost-efficiency for CX teams today:

  • Sentiment analysis to understand if a customer is upset, happy, etc. and to prepare agents to engage accordingly with the customer 
  • Suggested responses to customer issues on calls, email, text, and social, delivered to agents in the company’s brand voice and tone 
  • Intelligent search to pull up relevant information in your knowledge base or contact center platform faster
  • Intelligent routing to send calls and tasks to the right customer service agents, based on the customer’s specific needs, language, or even sentiment
  • Call summaries to eliminate the tedious work of note taking for agents, while still providing detailed resolution and issue tracking  
  • Multilingual support by translating articles, emails, product pages, and materials to assist customers in their native languages 
  • Smarter chatbots that can detect user’s issue, language, sentiment, and more, to offer the best possible response in a conversational way 
  • Interactive Voice Response (IVR) systems to direct customers by sharing their reason for calling in a conversational way 
  • Powerful insights based on all the data you’re collecting to understand customers, agents, performance, trends, opportunities, and more

While there are press releases announcing the launch of a new AI solution every week, the truth is that many of them are being built as we speak. 

All the CX and CRM software solutions know they need to be in the AI game as it becomes a bigger part of the future of customer experience. Understanding what to look for, how to get buy-in internally, and how to implement AI properly, are all part of the process that CX leaders will need to learn to be successful in this new AI world. 


AI lingo got you confused? Check out our AI Glossary for common terminology.