The AI Customer Service-Empathy Disconnect
I recently came across a video that perfectly captures the massive disconnect between what corporate executives care about and what human beings actually need.
In the video, a woman in a wheelchair was sharing her experience of being illegally discriminated against and denied access to a bus. But what caught my attention wasn’t just the awful situation itself; it was how she described her interaction with a customer representative from that very same company.
She actually shouted the rep out because of how deeply human their reaction was. The representative didn’t just read a corporate script about “repercussions” and “unacceptable behavior.” The rep was so genuinely furious and devastated on behalf of the woman that the woman described the rep as being “hysterical.”
Think about that. A customer support agent was so moved by a customer’s mistreatment that they lost their corporate filter and got emotional.
That is raw, empathetic, human-to-human solidarity, and it is something AI can never replicate.
The AI Backlash is Already Here
We are reaching a tipping point where people are getting fiercely protective of human spaces. If you spend any time on storytelling YouTube channels, you’ll notice a massive uptick in comments picking apart whether the audio is narrated by a person or an AI. When it’s AI, the feedback is overwhelmingly negative.
Sure, for a sentence or two, synthetic voices can trick us. But once you get past a paragraph, most people sense that something is “off.” The unfortunate side effect is a hyper-paranoid culture where real voice actors, video creators, and graphic designers are constantly being accused of using AI just because the tech is getting better.
But “better” doesn’t mean perfect, and it certainly doesn’t belong in customer service, no matter how much executives want to force it on us.
The Knowledge Base Trap
Look, I am not anti-AI. I am anti-bad AI. I’m anti-“shoving AI into literally everything with zero strategy, breaking systems that weren’t broken in the first place.”
When a chatbot is implemented thoroughly, it can be incredibly helpful. But there’s a fundamental flaw in how most companies build them, and it ties right back to core messaging. It has to do with the fact that Chatbots pull their knowledge from something called a knowledge base. A knowledge base is essentially a database of all of the rules of the company and how the bot is supposed to interact with customers regarding a wide variety of comments and complaints. But there’s an issue with knowledge bases:
The Flaw: Written for insiders, not customers.
Knowledge bases are usually written for people who already know how the system works. They are packed with internal terms, technical jargon, and structural assumptions. But when a customer reaches out, they don’t know your code, they don’t know your UI history, and frankly, it is not their job to know.
They should be able to log in, solve their problem, and leave. If they wanted to understand the structural complexities of your software, they’d build a cheaper version themselves.
When you trap a customer in an AI loop, you force them to guess the exact right “magic words” to trigger a helpful response. If they don’t know the jargon, they get hit with repeated, robotic phrases. It’s incredibly frustrating.
Short-Term Margins vs. Long-Term Health
Let’s stop giving these companies the benefit of the doubt. Executives aren’t rushing to implement AI because they mistakenly think it improves the user experience. They know customers don’t want it. They do it anyway solely for the profit margins. They look at AI and see a golden ticket to cut staff, slash overhead, and pad their own pockets.
It is a short-sighted, exhausted strategy. They are trading long-term business health and customer loyalty for immediate, artificial growth to please investors and look “future-forward.”
It’s strikingly similar to the early days of poorly managed offshoring. Companies outsourced customer support to cheaper regions to save a buck, but because they refused to train the reps properly, it just created a legacy of customer friction. Half-baked AI is doing the exact same thing, just with code instead of a phone line.
AI can be brilliant when it genuinely adds value. But right now, a lot of companies are willingly sacrificing the exact thing that keeps customers around—actual, empathetic human connection—all to save a dollar. I really wish they’d stop.
Want to make your business more human?
Let’s move past the scripts and the generic chat loops. Connect with us to discover how to design customer experiences that people actually love.
Let’s Chat