AI is rapidly automating call centers, aiming for full replacement, not just efficiency. While AI excels at data processing, human insight remains crucial. Leaders still need nuanced judgment and experiential knowledge that current AI tools lack for effective decision-making, despite early automation successes.
SwissCognitive Guest Blogger: Terryel Hu – “AI is Winning Against Call Centers. Yet Human Drivers Remain Essential”
While the initial hype may be stabilising, overall investment in AI remains high. Recent data from PitchBook revealed that AI and machine learning startups received $73.1 billion in venture capital funding, making up 57.9% of total investments. This remains a massive investment by global standards.
Like all emerging technologies, the longer-term benefits of AI might not yet be fully realised. But there are some quick wins worth noting. The clearest example? Call centres. AI agents have largely been deployed to automate predictable, repetitive jobs such as call centres, admin, and customer support roles.
The banking sector is a case in point. In highly regulated environments such as Australia, the larger banks have still managed to reach a major milestone. Commonwealth Bank has begun trialling AI chatbots to replace call centres. ANZ has recognised the potential to fast-track how it verifies loan contracts. Similar examples can be found across the global banking sector. This marks a significant shift.
Chatbots Aren’t Speeding Up Call Centres. They’re Replacing Them.
The move to implement AI chatbots is just one part of the automation strategy. Chief Technology Officers are familiar with Robotic Process Automation, while business leaders are well-versed in outsourcing and acquisitions. As costs continued to rise, initiatives for ‘doing things faster’ were already on the table. The sudden release of Agentic AI has only accelerated that need.
But here’s the catch: the ultimate aim of AI was never to make things more efficient—it was about eliminating them entirely. Call centres and their traditional management methods are both at risk of being replaced. This shift challenges the value of incremental improvements offered by classic efficiency playbooks like Lean, Six Sigma, and Agile. If replacing call centres is now the standard, small efficiencies may no longer be the game.
Can AI Produce Knowledge that Leads to a Competitive Advantage?
Replacing call centres is a promising experiment. It shows companies what can be automated and what still requires human judgment. Yet most current AI tools don’t generate a true competitive edge. Why? Because chatbots process information, but rarely produce actionable knowledge.
Researchers distinguish between data, information, and knowledge—a complex debate among cognitive scientists. For simplicity:
Data is raw (numbers, words). Data is organized into information. When we attempt to add meaning to information, it becomes knowledge.
AI does produce some form of knowledge. Most AI companies are deploying chatbots—ChatGPT-style tools—that produce declarative knowledge (factual statements), not procedural knowledge (how-to processes). It’s not the kind of knowledge humans use to ride a bike, cook a dish, or lead a team.
Leaders require knowledge that involves judgment and experience. For example: How to predict customer churn more precisely? How to improve medical diagnosis? How to reduce excavation wastage in mining? The analytical facts produced by generative AI are often missing the business acumen that only human leaders possess. That’s not a design flaw nor a technical fault. As the saying goes, “There is no replacement for experience.”
No matter how clever your prompt, generative AI tools solve only clearly defined problems. Seasoned leaders still have to ask, “What’s my purpose?” and likewise, their executives may wonder, “What’s my mission in life?” These aren’t queries with clean answers—they require human judgment that’s hard to replicate.
Can AI Agents Learn Like Us?
A common misconception is that humans learn like machines—through repetition, identifying errors, and iterating. Language models may understand the mathematical definitions of error and solution, but humans bring meaning to these processes. For example, a bot might be trained to view smoking as an error. A user may agree—but still fail to quit despite trying every method. The reasons behind their behaviour are personal. Maybe the solution hasn’t been discovered yet.
A chatbot may provide a smooth conversation, but it can’t tell whether something is an error for you. That’s a personal insight. Learning is not just repetition, one trial after another. It’s also the accumulation of experience, reflection, and adjustment.
AI’s Call-Centre Experiment Is a Good Win. It’s Time to Focus on the Next One
The ChatGPT style of AI has been rapidly embraced across countries, cultures, and sectors. Remarkably, OpenAI’s ChatGPT reached 100 million users just two months after launch—making it one of the fastest adoptions in digital history.
One of the clearest examples of its usefulness is in replacing call centres and support roles. This shift aligns well with the current capabilities of AI. The move to automate call centres has opened new doors for industries that were previously manual and process-heavy. For larger organisations, something as standard as a chatbot may be enough. Banks certainly think so. They’ve shown how chatbots can serve wide-ranging functions—from loan processing to reliable customer service.
A greater challenge—and opportunity—lies in finding how AI can support larger, more complex decisions and transactions. Only by experimenting and staying alert to AI’s impact can companies discover where its true advantage lies.
About the Author:
Terryel Hu is a founder, advisor, and innovator recognised by Thinkers360 as a Top 50 Leader in Management. He is dedicated to helping businesses bridge capability gaps and raise awareness about AI strategy.
AI is rapidly automating call centers, aiming for full replacement, not just efficiency. While AI excels at data processing, human insight remains crucial. Leaders still need nuanced judgment and experiential knowledge that current AI tools lack for effective decision-making, despite early automation successes.
SwissCognitive Guest Blogger: Terryel Hu – “AI is Winning Against Call Centers. Yet Human Drivers Remain Essential”
Like all emerging technologies, the longer-term benefits of AI might not yet be fully realised. But there are some quick wins worth noting. The clearest example? Call centres. AI agents have largely been deployed to automate predictable, repetitive jobs such as call centres, admin, and customer support roles.
The banking sector is a case in point. In highly regulated environments such as Australia, the larger banks have still managed to reach a major milestone. Commonwealth Bank has begun trialling AI chatbots to replace call centres. ANZ has recognised the potential to fast-track how it verifies loan contracts. Similar examples can be found across the global banking sector. This marks a significant shift.
Chatbots Aren’t Speeding Up Call Centres. They’re Replacing Them.
The move to implement AI chatbots is just one part of the automation strategy. Chief Technology Officers are familiar with Robotic Process Automation, while business leaders are well-versed in outsourcing and acquisitions. As costs continued to rise, initiatives for ‘doing things faster’ were already on the table. The sudden release of Agentic AI has only accelerated that need.
But here’s the catch: the ultimate aim of AI was never to make things more efficient—it was about eliminating them entirely. Call centres and their traditional management methods are both at risk of being replaced. This shift challenges the value of incremental improvements offered by classic efficiency playbooks like Lean, Six Sigma, and Agile. If replacing call centres is now the standard, small efficiencies may no longer be the game.
Can AI Produce Knowledge that Leads to a Competitive Advantage?
Replacing call centres is a promising experiment. It shows companies what can be automated and what still requires human judgment. Yet most current AI tools don’t generate a true competitive edge. Why? Because chatbots process information, but rarely produce actionable knowledge.
Researchers distinguish between data, information, and knowledge—a complex debate among cognitive scientists. For simplicity:
Data is raw (numbers, words). Data is organized into information. When we attempt to add meaning to information, it becomes knowledge.
AI does produce some form of knowledge. Most AI companies are deploying chatbots—ChatGPT-style tools—that produce declarative knowledge (factual statements), not procedural knowledge (how-to processes). It’s not the kind of knowledge humans use to ride a bike, cook a dish, or lead a team.
Leaders require knowledge that involves judgment and experience. For example: How to predict customer churn more precisely? How to improve medical diagnosis? How to reduce excavation wastage in mining? The analytical facts produced by generative AI are often missing the business acumen that only human leaders possess. That’s not a design flaw nor a technical fault. As the saying goes, “There is no replacement for experience.”
No matter how clever your prompt, generative AI tools solve only clearly defined problems. Seasoned leaders still have to ask, “What’s my purpose?” and likewise, their executives may wonder, “What’s my mission in life?” These aren’t queries with clean answers—they require human judgment that’s hard to replicate.
Can AI Agents Learn Like Us?
A common misconception is that humans learn like machines—through repetition, identifying errors, and iterating. Language models may understand the mathematical definitions of error and solution, but humans bring meaning to these processes. For example, a bot might be trained to view smoking as an error. A user may agree—but still fail to quit despite trying every method. The reasons behind their behaviour are personal. Maybe the solution hasn’t been discovered yet.
A chatbot may provide a smooth conversation, but it can’t tell whether something is an error for you. That’s a personal insight. Learning is not just repetition, one trial after another. It’s also the accumulation of experience, reflection, and adjustment.
AI’s Call-Centre Experiment Is a Good Win. It’s Time to Focus on the Next One
The ChatGPT style of AI has been rapidly embraced across countries, cultures, and sectors. Remarkably, OpenAI’s ChatGPT reached 100 million users just two months after launch—making it one of the fastest adoptions in digital history.
One of the clearest examples of its usefulness is in replacing call centres and support roles. This shift aligns well with the current capabilities of AI. The move to automate call centres has opened new doors for industries that were previously manual and process-heavy. For larger organisations, something as standard as a chatbot may be enough. Banks certainly think so. They’ve shown how chatbots can serve wide-ranging functions—from loan processing to reliable customer service.
A greater challenge—and opportunity—lies in finding how AI can support larger, more complex decisions and transactions. Only by experimenting and staying alert to AI’s impact can companies discover where its true advantage lies.
About the Author:
Terryel Hu is a founder, advisor, and innovator recognised by Thinkers360 as a Top 50 Leader in Management. He is dedicated to helping businesses bridge capability gaps and raise awareness about AI strategy.
Share this: