The more important question might be: why are we still struggling with the same training problems we've had for twenty years?

In this episode, we tackle a challenge facing nearly every contact center: new hire training that produces high attrition, slow ramp times, and inconsistent performance.

The conversation started with a client losing roughly half of its new hires during training and another portion during nesting. Despite new technology, the outcomes felt painfully familiar.

We explored some uncomfortable questions:

• Are organizations hiring the right people for the job?

• Do recruiters actually understand the roles they're filling?

• Are training programs teaching agents what to know instead of how to find answers?

• Why are contact centers still trying to memorize information that changes every few weeks?

• How much time do leaders spend observing their own training programs?

Bob argues that many training failures begin before day one, with hiring processes that prioritize filling seats instead of finding the right fit.

We also discuss one of the biggest missed opportunities in modern training:

AI is changing how agents work, but many training organizations haven't changed how they train.

As AI-powered knowledge systems, agent assist tools, and automation become standard, training leaders need a seat at the table. Yet in many organizations, training teams are disconnected from the technology decisions that will fundamentally reshape agent performance.

One of the biggest insights from the discussion:

The real opportunity isn't using AI to automate training.

It's using AI to automate the things training used to spend time on so trainers can focus on the skills that matter most:

• Building rapport

• Problem solving

• Judgment

• De-escalation

• Relationship-building

• Handling difficult conversations

Technology changes.

The fundamentals don't.

Topics discussed:

• Why new hire attrition remains so high

• Hiring mistakes that create training failures

• Teaching agents how to find answers versus memorizing answers

• The role of nesting and floor support

• Why training content quickly becomes obsolete

• The disconnect between training teams and AI initiatives

• Agent assist and the future of onboarding

• Tough skills versus technical skills

• Why fundamentals still matter in modern contact centers

• How AI should reshape training priorities

Bob just returned from Italy with a story that should make every customer service leader pay attention.

At train stations in Venice and Florence, there were no employees to help. Just kiosks. If you wanted a ticket, you figured it out yourself. If you had a question, there was nobody to ask. It wasn't a glimpse of the future. It was the present.

That experience led us into a bigger discussion about AI, automation, and what customer service becomes when human interaction disappears.

We unpacked a recent Anthropic report showing that customer service roles have some of the highest exposure to AI-driven task automation. But exposure to tasks is not the same as elimination of jobs.

The deeper question is this:

Is customer service simply a collection of transactions, or is it fundamentally about relationships?

We discussed real-world results from an enterprise deployment of agentic AI where:

• Escalation rates were 4x higher when customers interacted with AI versus humans.
• Customers were significantly more likely to demand supervisors from bots.
• Contact volume increased by 50% in less than six months.
• Companies discovered that delivering bad news remains far more effective when done by a human.

History suggests that new channels rarely reduce demand. Email didn't reduce contacts. ATMs didn't eliminate bank tellers. They changed the nature of the work.

AI may do the same.

At the same time, organizations are racing toward automation while learning that token costs, increased interactions, and customer behavior may complicate the promised economics.

The technology is arriving at bullet-train speed.

The question is no longer whether AI is coming.

The question is:

Who are you in an AI-first world?

Will your company become a vending machine that happens to sell products?

Or will you intentionally preserve the human elements that create trust, loyalty, and relationships?

Because customer relationship management was never supposed to become customer technology management.

Topics discussed:

• Anthropic's AI exposure findings
• Why task automation doesn't automatically eliminate jobs
• The difference between transactional and relational service
• Real-world lessons from agentic AI deployments
• Rising escalation rates with AI interactions
• The hidden cost of token consumption
• Why customers treat bots differently than humans
• The future role of human agents
• How leaders should rethink customer service strategy in an AI-first era

AI hype is colliding with operational reality. A shoe company gains $127M in value by saying “AI,” while contact center leaders are told their entire model is obsolete. The shift from CCaaS to “Customer Experience Automation” reframes everything: not just support, but marketing, sales, and service collapsing into one AI-driven layer.

The problem: the foundation is broken. Knowledge is fragmented. Customer data is duplicated. Organizations are misaligned.

This episode dissects the gap between what the industry is promising and what companies can actually execute—and why customer service leaders are about to become the last line of defense when it fails.

Key Quotes

• “Customer experience automation just means AI is now the one doing the talking.”
• “Your human agents can’t find the right answers today—but now the AI is supposed to?”
• “This isn’t a technology problem. It’s an organizational problem.”
• “If this fails, customer service cleans it up. Again.”
• “The train is moving. You either help steer it or get run over by it.”

Practical Takeaways

• Stop debating AI capability. Start fixing knowledge.
• Treat data quality as a blocking issue, not a backlog item.
• Force alignment between marketing, sales, and service before automation.
• Assume AI will act autonomously—and design safeguards accordingly.
• Position customer service as the control layer, not the endpoint.