Human-in-the-Loop as Strategy
Many, if not most people, simply don’t trust AI. Of course, part of that is caused by the fact they don’t yet really know AI, but they have you to help change that. Right?
Please Note: Once again this is a particularly long post, so subscribers will find the emailed version of this issue cut off at the end. Please go to the web version for the complete version.
If You Find AgenticMSP Valuable: Please recommend AgenticMSP to your friends and colleagues. Also, please offer up your comments, observations, questions, or any other responses in the chat below. Let’s get a hearty conversation going!
One issue you’re very likely facing in your customer conversations about AI is the issue of safety. Customers are concerned about core issues with AI such as security, data privacy, governance, regulatory compliance and, most importantly, trust. Many, if not most people, simply don’t trust AI. Of course, part of that is caused by the fact they don’t yet really know AI, but they have you to help change that.
The best way to assure trust is to maintain vigilance. The farmer says their shadow is the best fertilizer. We’re told we should trust but verify. Then ZTNA taught us to trust nothing, verify everything.
With AI, the best way to preserve trust is to keep a human-in-the-loop (HITL) at all times. Supervised AI agents check in with a human on a regular, scheduled basis to assure they remain on the right track. Requiring human approval to transactional or any other commitment-bearing task protects the organization from inadvertent errors.
The End is at The End – But it’s Just the Beginning
The last news article in today’s AgenticMSP is unusual in that it’s one of my own. It is the last article I wrote for my Human-in-the-Loop column in Application Development Trends (ADT) magazine.
This column began in October 2024 as “The Citizen Developer” which focused on low-code/no-code (LCNC) platforms that allowed users to build applications by moving tiles around a screen to define processes and workflows.
It was less than six months before we realized LCNC was already on its way out, replaced by AI agent builders users could just tell what they wanted. The agent would then build the requested software. They obsoleted LCNC in truly record time.
Unfortunately, for various reasons beyond my or my editor’s control, the HITL column has been suspended.
The idea of keeping a human-in-the-loop is just too important a priority to simply let that be that, so you’ll see me moving more articles about HITL here in AgenticMSP.
Help Me Out Here
As with everything we discuss here, I’m hoping you’ll enthusiastically share with me your stories about how keeping a human-in-the-loop helped you achieve great AI success. When something about HITL strikes you as important, please share it in the chat!!
And now, the news:
