The AI Winter is Coming: Will Chatbots Survive the Next Tech Cycle?

AI chatbots have made a significant impact on customer service, particularly in contact centers. With some experts predicting a slowdown in AI investments, there's uncertainty about whether AI chatbots will continue to thrive or if businesses might revert to human-only agents to deliver more personalized service. 

According to Gershon Goren, Founder + CEO of Cangrade, the demise of chatbots is far from inevitable: “While not all pollyannaish predictions for Generative AI are coming true yet, the demise of chatbots is not in the cards. Improvements in GenAI technology by the largest players—Nvidia, OpenAI, Anthropic, and others—combined with strengthening interest from other tech companies…are powering a new wave of AI chatbots.”

In this Q&A with Bobby Hakimi, Chief Product Officer and Co-founder of Convoso, we explore how these advances are shaping the future of AI chatbots in contact centers and the practical steps companies can take to harness their potential effectively.

1. Many experts predict an “AI Winter” where investment in AI will slow. What makes you confident that AI chatbots will thrive despite this trend, especially in contact centers? Could we see a shift where companies revert to human-only agents to enhance personalization?

The trend for using AI in contact centers is actually building. It’s not going to slow down. We’ve just scratched the surface. AI will become the front line of communication. As it becomes smarter, it will be able to handle personalization better. AI is really a must-have efficiency tool that’s helping streamline operations. So, contact centers will continue to adopt AI. It will continue to be implemented in more parts of technology and people's everyday lives.

But it takes thinking through how AI fits into your specific company. It takes strategy and training for the AI to engage with more personalization. I think once contact centers understand the improvements to efficiency and the cost savings, more will adopt the technology. The trend is actually that you need AI to stay competitive.

2. While AI chatbots can reduce agent burnout and handle repetitive tasks, some critics argue they dilute the human touch in customer service. How do you respond to concerns that relying too much on AI could harm customer relationships in the long term?

I think it will be the opposite at some point. I think customers are going to prefer to talk to AI. AI will be faster because it has immediate access to vast amounts of information and can resolve issues more efficiently than a human who might lack the necessary training or knowledge. Unlike humans, AI doesn’t have off days or emotional fluctuations, ensuring a consistent and reliable service experience every time. People will increasingly prefer interacting with AI due to its speed, efficiency, and consistent performance as time passes.

However, it depends on the customer’s needs. Like all technologies, AI chatbots are built for specific purposes and, therefore, have limitations. At this point, it functions best for those data-based responses to quickly solve an issue or qualify caller needs or initial questions. But then they can be routed to a human agent if the AI hasn’t resolved their call reason. Most people are trying to solve something, so if the AI helps take care of that efficiently, then it’s actually helping to improve the customer experience. A human is there to give that personalized interaction when necessary, so for now, this combination can actually help to enhance customer relationships. These are just different ways of meeting the customers' needs effectively. 

3. As AI chatbots evolve to handle more complex conversations, is there a risk that businesses become too dependent on AI and neglect the development of human agents? How do you strike the right balance between AI and human expertise?

It’s a false choice. Asking about becoming too dependent on AI is like asking if we will depend on the Internet to send emails or continue only using the post office. 

This is just what’s happening. If you don't adapt, you will fall behind.

We need to let humans do more complex problem-solving rather than sharing information and answering questions on the phone about stuff that a robot can handle. With AI taking up tedious and repetitive tasks, more resources and time can be allotted to human agents to handle the more engagements best served by person-to-person interactions. 

4. What do you say to contact centers hesitant to adopt AI because they fear the technology may stagnate or become obsolete in the next tech cycle? How do you future-proof AI solutions in a field that’s constantly changing?

Use one of the larger vendors, like Google or Open AI, because they'll keep evolving. You’ll never see Google not keeping up with the times. So, use a vendor that’s constantly innovating. Don't use a vendor that is building their own. At Convoso, for example, we don't build our AI – we use open AI. If you’re a contact center, use one of these major vendors to keep up with the functionalities and inherit all the new features it comes with.

The nature of AI with deep learning models is to constantly adapt and improve. AI is not a set-it-and-forget technology that you just buy and turn on. It needs to be set up correctly in the first place and then maintained. If not, I suppose it could seem obsolete, but the reality is that’s what you did by not training your AI properly, not the technology itself. It’s the same thing with humans - If you don't keep training agents, they will stagnate and not keep up.

5. The narrative often focuses on how AI benefits businesses, but what are the lesser-known challenges companies face when implementing AI chatbots? What are some common misconceptions about the “ease” of AI integration in contact centers?

There are some misconceptions about what it takes to train the chatbot effectively. Just like training human employees, if they have the information but don’t get the proper direction, guidance, and mentorship, then they may make mistakes. It’s the same with AI and GPT. You need to let it know exactly what you want it to do. The more specific the prompt is, the more powerful and quality the output will be.

AI is not a plug-and-play tool that just runs your campaigns and transforms your outcomes. It takes careful setup, and you must continually refine the AI’s ability to understand customer intent. You have to dedicate time and the right level of expertise. 

Using AI tools requires customizations to fit the contact center’s specific needs, such as deep integration with an existing CRM and other backend systems. This process can be more complicated than people expect, and they can get frustrated because it involves trial and error to fine-tune the AI’s performance for different customer scenarios.

Balancing automation with human intervention is also challenging. While your AI can handle many tasks, contact centers still need a system to smoothly escalate more nuanced or emotional interactions with human agents. 

People sometimes think AI will deliver immediate ROI. But it takes time to optimize the system before you’ll see a return on investment. There’s basically a learning curve for both the AI and the company, so you have to think of it in terms of the long game.

6. Do you foresee a future where AI chatbots are capable of handling not only more complex conversations but also deeply emotional customer interactions? How close are we to AI reaching that level of sophistication, and should we even aim for it?

Yes, we’re definitely getting closer to up-leveling the complexity of conversations and perceptions of emotional tone. The AI will use sentiment analysis and be able to gauge the best response. In the future, it’s going to have empathy. The improvements are happening gradually, but we should see this in the next few years. 

7. Outbound sales and lead gen teams that don’t adopt AI may fall behind—but what about smaller businesses with limited budgets for AI? Are there effective ways for them to integrate AI chatbots without breaking the bank, or do they risk being left behind?

You don't need to have a big budget to do AI. It's very cheap these days.

So many options are available now, and that will continue, which will inevitably lead to more affordable solutions. There can be the expense of having someone on staff with the technical knowledge to set it up properly and maintain it. But these days, some YouTube videos teach you how to talk to AI and do prompting. It doesn't take a genius to figure it out. You need to have somebody intelligent and creative, and they can figure it out. I think people are making it a bigger deal than it is in terms of difficulty.

Takeaways

As AI chatbots become more sophisticated, businesses that thoughtfully integrate and manage this technology will likely benefit substantially.

As Goren notes, “Unlike its generic predecessors: ChatGPT, Claude, and Gemini, this new generation of chatbots are often built into existing products and services, enhanced by unique, industry-relevant data. While these chatbots have more limited applications in the markets they serve, they also provide greater value than generic chatbots, leading to more accurate and targeted responses for specific purposes. This creates a significantly rosier outlook for improved workforce productivity and a subsequently larger effect on the economy, all thanks to GenAI.”

Although there are challenges—such as finding the right balance between automation and human input—these advancements suggest that chatbots will continue to play a critical role in improving efficiency and productivity.

AI chatbots still have the potential to reshape customer interactions and deliver measurable results.

Subscribe to The CTO Club’s newsletter for more AI interviews and insights.