The Worst Possible Use of AI

AI is powerful when it speeds up internal work, removes repetitive admin, and helps teams move faster. But Gayle Rogers argues there is a line you should not cross: using AI to replace customer interaction as a default. He calls it “the absolute worst use of AI in the history of anything” because it turns the very thing that creates trust, human connection, into a maze.

This matters because most businesses don’t lose customers only to competitors. They lose customers to friction. People don’t have time to fight a system just to ask a question. And when they feel blocked, they attribute that feeling to the brand, not the tool.

When Automation Creates Friction

Automation is supposed to shorten steps. When it adds steps, it becomes a liability.

Rogers gives an example that every decision-maker should pay attention to. He called a local business to ask permission for a video shoot. The call was answered by an AI agent, “I think named Eric,” which immediately began a pitch and then demanded to know who he wanted to connect with without offering guidance.

The Six Minute Call Problem

The moment the call became difficult, the whole experience collapsed. Rogers says he had to repeat his name four times because the system kept responding with the same block: it needed his name before it could move forward. Then it asked for his phone number “in case we get disconnected” and repeated the number back one digit at a time. Eventually he reached a real receptionist, but the receptionist wasn’t even taking calls because the AI was acting as the gatekeeper. The end result was still voicemail. What should have been a two-minute call became a six-minute call.

This is the key lesson: bad AI doesn’t feel modern. It feels like the business is hiding.

Efficiency Without Connection Is a Trap

There is a difference between automation that helps a customer and automation that protects a company from customers. The first reduces friction. The second increases it.

When a system makes a customer repeat information, mishears them, or refuses to move forward, it communicates something you would never say out loud: your time matters less than our process. That perception damages trust, and trust is the foundation of conversion, retention, and referrals.

Businesses often justify this with efficiency. But efficiency that drives customers away is not efficiency. It’s cost-cutting disguised as innovation.

Where AI Helps and Where It Hurts

AI can be excellent in customer service when it supports humans rather than replacing them.

Good uses typically include:

• Routing and triage when it’s fast and accurate
• Drafting internal summaries for agents to respond faster
• Suggesting answers to common questions while a human stays accountable
• After-hours intake that promises a next-day callback with a real person

Bad uses look like Rogers’ example:

• AI as the primary gatekeeper to a human
• Systems that require repeated inputs to proceed
• Long verification scripts before the customer can explain what they need
• AI that forces voicemail anyway

Keep Humans in the Moments That Matter

A simple rule helps. If the customer’s intent is emotional, urgent, or ambiguous, keep a human in the loop. If the customer is trying to do something simple: book, ask, confirm, then automation must make it faster, not slower. If it can’t, remove it.

A Simple Audit to Fix the Experience

If you want to avoid becoming the business people complain about, run a quick audit this week:

1. Call your own number from a mobile phone in a noisy environment and time how long it takes to reach a human.
2. Track how many times you’re asked to repeat name, number, or intent.
3. Ask: did the system make the experience easier, or did it block the customer?
4. Identify the top five customer intents and ensure each resolves in under two minutes.
5. Add an immediate escape hatch: “Press 0 to speak to a person.”
6. Train your humans to pick up where automation stops, not restart the interrogation.

Rogers’ advice is direct for a reason. Technology should support communication, not stand between you and the people who pay you.