On the Verge of a Fourth Paradigm Shift In How I Work With AI

A note: this is about my personal cognitive relationship with the technology, not its application in finance or organizational transformation. That is a different conversation.

When I first started using ChatGPT in early 2023, I came with a problem-solving mindset. I had identified a problem, I had a solution in mind, and my job was to get the AI to produce it. Sometimes it worked. More often than not, the output fell short of what I had envisioned.

As the capabilities of both ChatGPT and Claude matured, I adjusted my approach. Rather than directing, I started collaborating. Providing proper context, introducing multiple angles, co-creating rather than prescribing. The shift was subtle, but it changed the quality of what I produced.

Over the past months, working with open-source agents has pushed me further. I stopped focusing on the problem in front of me and started focusing on the outcome I was trying to reach. What is the goal, not just what is the challenge? That reorientation expanded how I use curiosity as a working tool, not as a soft skill, but as a genuine driver of how I structure my thinking.

This week, an article opened the door to a fourth shift. Efficiency and effectiveness were the previous shifts. This one is about agency.

The prompt suggestions that resonated most:

  • Ask for hints, background thinking, or clarification, not for answers.

  • Start with a blank page. Write your own analysis first. Then ask AI to challenge your thinking, not produce it.

  • Ask for thinkers, not thinking: who has already addressed this problem, and what would they say?

  • Draw a clear line between rote work and creative work. Let AI handle functional emails. Keep your essays and memos yours.

  • Treat AI less like an encyclopedia and more like a personal trainer, something that builds your mental muscles rather than replacing them.

What I find most interesting about this fourth shift is the direction of travel. Each previous phase made me more capable with the tool. This one brings something different: How do I ensure the thinking remains mine and keeps getting sharper? 

Where are you in your own AI paradigm? Has your approach shifted, or stayed the same?

Annette van Berge Henegouwen


Website: https://www.annettevanbergehenegouwen.com/

Next
Next

Learning Outperforms Control. But Only When The System Allows It