Experience can't be prompted
Updated: Aug 15
I had an interesting conversation this week with a few fellow fractional CIOs and CTOs that genuinely made me stop and think.
A couple shared situations where business leaders had recently compared their advice and recommendations against a quick ChatGPT or Claude response. Not a detailed prompt with business context, constraints, risks and background. Just a short question typed into AI.
That was a bit of a lightbulb moment for me.
I’m a huge supporter of AI and use it every day. I absolutely believe it can improve productivity, generate ideas and accelerate decision making. But there’s a growing risk that we end up placing too much confidence in AI responses without enough human judgement around them.
Experienced technology leaders like CTOs are rarely just answering the question directly in front of them. They’re drawing on years, sometimes decades, of lived experience. Past successes and failures. Commercial awareness. Risk understanding. Organisational culture. Timing. Stakeholder dynamics.
Things that don’t appear in a generative AI response to a short prompt.
AI can absolutely provide answers. But judgement and experience is something different.
It also made me reflect on my recent post around whether we may slowly be outsourcing parts of our critical thinking as AI becomes embedded into daily work.
Perhaps one of the unintended consequences of AI adoption isn’t just the erosion of critical thinking over time, but also the gradual undervaluing of deep human expertise because AI can produce fast, confident sounding responses in seconds.
Responsible AI governance isn’t just about security, regulation and compliance. It’s also about continuing to value human experience, informed challenge and critical thinking alongside the technology.
That feels like something worth reflecting on. Your thoughts are very welcome.
Comments