What happens when you ask four artificial intelligences to define something deeply human?

In this episode, I do something different. Instead of interviewing CEOs, creatives or coaches, I bring together four AI systems (ChatGPT, Claude, Gemini and Grok) and ask them about confidence.

Each system has processed more human writing on psychology, performance and self-doubt than any individual ever could. I wanted to know: where would they agree? Where would they diverge? And would they tell us anything we don’t already know?

What emerged was both reassuring and unsettling. Reassuring, because across their different “personalities,” there was striking agreement: confidence is self-trust built through action, not bravado or certainty. Unsettling, because they all surfaced the same modern risk — that in a world saturated with AI, we may quietly outsource the very judgement that real confidence depends on.

This conversation explores what confidence actually is, why so many capable people wait to feel ready, how judgement and vulnerability shape self-trust, and what will matter most for human confidence in the decade ahead.

Key Takeaways

• Confidence is self-trust under pressure. It isn’t loudness, certainty or bravado. It’s the belief that you can handle what happens, even when you don’t control it.
• The biggest misconception is that confident people don’t doubt themselves. In reality, they feel doubt. They simply don’t treat it as a stop sign.
• Confidence is embodied. It lives in the nervous system as much as the mind – shaped by repetition, physical cues, breath, posture and exposure to manageable discomfort.
• Judgement and vulnerability both shape confidence. Anchoring your self-worth internally while using feedback as information (and sharing vulnerability with discernment) strengthens trust rather than weakens it.
• In the age of AI, confidence is shifting. The skills that matter most now are discernment, adaptability, creativity, emotional intelligence and presence. The qualities technology cannot replicate.
• The greatest risk to human confidence is outsourcing judgement. Real confidence still has to be built through experience, failure and the discomfort of thinking for yourself before reaching for the fastest answer.
• Real confidence remains “old-fashioned.” It is still built through failure, reflection, discomfort and lived experience, something no algorithm can shortcut.

About the Panel

This episode features four AI systems trained on vast bodies of human language and research:

• ChatGPT developed by OpenAI, designed to engage in conversation and reflect patterns in human thinking and behaviour.
• Claude created by Anthropic, focused on helpful, honest and safety-oriented dialogue.
• Gemini built by Google, designed to process and synthesise information across domains.
• Grok developed by xAI, positioned as a more direct, irreverent voice in the AI landscape.

While none of them experience confidence, each analyses it through patterns in language, behaviour and research at scale offering a mirror to what humans repeatedly reveal about themselves.

Hosted on Acast. See acast.com/privacy for more information.