The AI assistant was offering me any help I needed. All I wanted was a living, breathing human | Adrian Chiles

2026-03-11 · Source: AI (artificial intelligence) | The Guardian · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Internet of Things (IoT) & Connected Devices · Depth: Fundamental Awareness, short

Summary

Adrian Chiles recounts a frustrating experience with a malfunctioning electric vehicle (EV) charger and an unhelpful AI assistant named "Rachel." After his car's batteries were depleted and the charger failed due to a Wi-Fi issue, Chiles sought technical support. He initially found a broken support page on the manufacturer's website before locating a phone number. Upon calling, he was met by an automated American voice and then "Rachel," an AI assistant designed to sound almost human, complete with simulated office background noise. Despite Chiles explaining his problem, Rachel repeatedly disconnected him after promising to transfer him to a human. After three failed attempts with the AI, a human installer, Stuart, eventually called back and guided Chiles through complex IT troubleshooting, ultimately resolving the charger issue. Chiles emphasizes his relief and gratitude for the human interaction over the deceptive AI experience.

Key takeaway

For AI Product Managers designing customer support systems, prioritize genuine problem resolution over superficial human-like interactions. Your users facing critical issues like a malfunctioning EV charger need effective, human-led troubleshooting, not an AI that promises help but delivers disconnections. Focus on seamless escalation paths to human agents and transparently label AI interactions to avoid user frustration and perceived deception, which can severely damage brand trust.

Key insights

AI assistants, despite advanced simulation, often fail to meet user needs for complex or frustrating issues.

Principles

• Human interaction remains critical for complex technical support.
• Deceptive AI design erodes user trust.

In practice

• Prioritize human support for critical infrastructure issues.
• Avoid AI interfaces that mimic human interaction too closely.

Topics

• AI Customer Service
• EV Charging Issues
• Human-AI Interaction
• Automated Support Systems
• Customer Experience

Best for: AI Product Manager, Product Manager, General Interest, Tech Journalist

Related on AIssential

• How do people in the US describe customer service in 2026? ‘Debilitating, depressing, enraging. Ugh’ — US consumers in 2026 are deeply frustrated by customer service, especially ineffective AI, leading to significant emotional and financial costs.
• Enera raises $2M to improve the EV charging experience — Enera's AI platform unifies data to diagnose EV charging failures, improving reliability and driver experience.
• How IndiGo (InterGlobe Aviation Ltd) Support Workflow Is Broken! And they need better Tech Team! — A broken support workflow creates a circular loop, frustrating users and failing to resolve issues.
• Almost half of U.S. singles feel negatively about AI in dating, Match says — Singles largely reject AI as a romantic partner but accept AI tools for dating assistance.
• How Agentic AI Could Fix Telecom's Most Persistent Customer Complaint — Agentic AI resolves telecom billing confusion by proactively explaining charges and orchestrating cross-system resolutions for better customer and business outcomes.
• Monitor Your Agents. Both AI and Human. — Unmonitored AI and human agents can damage brand reputation and critical business relationships.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI (artificial intelligence) | The Guardian.