Summary & Key Takeaways ​

• Jon Udell critiques the common phrase "human in the loop" for ceding authority to machines.
• He proposes reframing the concept to "our loop," where humans maintain agency.
• AI agents are then invited into this human-controlled process as assistants.
• This perspective applies to agentic software development, avoiding black-box scenarios.
• The goal is agent-assisted processes, not human exclusion from a machine-driven loop.

Our Commentary ​

This really resonates with me. The language we use around AI matters so much, and 'human in the loop' always felt a bit... subservient. I love the idea of flipping the script, asserting that it's our process, and we're just bringing in some very smart tools. It feels more empowering, less like we're just babysitting an autonomous system. It's a subtle but powerful mental shift for how we approach agentic development.