The More You Trust AI, The Less You Think

Research shows confidence in AI suppresses the very scrutiny it most requires, and boredom is one of the last defences left.

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://open.substack.com/pub/johanosteyn/p/the-more-you-trust-ai-the-less-you?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/

There is a quiet irony at the heart of the AI-enabled workplace. The better these tools become, and the more we come to trust them, the less inclined we are to question what they produce. That instinct feels efficient. It is also the beginning of a measurable decline in the very capacities that make senior professionals worth their salaries, and it is time we treated it as a business risk rather than a personal habit.

CONTEXT AND BACKGROUND

The evidence is now specific. A study by Microsoft Research and Carnegie Mellon University, surveying 319 knowledge workers across nearly a thousand real-world tasks, found that higher confidence in a generative AI tool is associated with less critical thinking, while higher confidence in one’s own ability is associated with more. The concern is not hypothetical or distant. As reported in an authoritative global assessment, three months after AI diagnostic support was introduced, clinicians’ ability to detect tumours without assistance had fallen by six percent. Skill, it turns out, is not permanent. It is maintained through use, and it fades through disuse.

INSIGHT AND ANALYSIS

This is where an argument I made previously becomes sharper. I have previously written about why we still need boredom in an age of relentless AI stimulation, arguing that the unstimulated mind is where attention recovers and original thought takes shape. The cognitive science has moved on since then, and it now points to a mechanism. When an AI tool offers a fast, fluent and polished answer, it nudges the mind out of slow, deliberate reasoning and into fast, intuitive acceptance, and the erosion of independent judgement accelerates beyond certain thresholds of use. The trap is self-reinforcing. As the tool earns our trust, our scrutiny relaxes, which is precisely the moment when scrutiny matters most. Researchers describe this as cognitive offloading, the delegation of mental work to an external system, which frees up effort in the short term but weakens the underlying faculty over time.

IMPLICATIONS

For a board, this reframes AI adoption entirely. The risk is not only that a model produces a flawed output, but that the organisation slowly loses the collective capacity to notice. When strategic analysis is routinely outsourced to a machine, institutional reasoning and independent foresight quietly atrophy, leaving leadership less able to challenge the very systems it depends on. The answer is not to abandon these tools, which deliver real value, but to build deliberate friction back into how they are used. That means protecting time for unassisted analysis, requiring a first draft of thinking before a prompt is written, and treating moments of difficulty and boredom not as inefficiencies to be optimised away but as the exercise that keeps judgement fit. Cognitive fitness, like any other, is lost when the effort is removed.

CLOSING TAKEAWAY

The most capable leaders of the next decade will not be those who trust AI the most, nor those who refuse it, but those who know when to lean on it and when to switch it off and think for themselves. Confidence in a machine is not the same as competence, and the two are quietly pulling apart. The discipline worth cultivating, in ourselves and in the people we lead, is the willingness to stay in the difficult, unstimulated, effortful place a little longer, because that is where thinking actually happens. Trust the tool less, and think more.

Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net