After attending AIxPH in Baltimore, and relentlessly promoting the concept of trust and how it works in consumer environments in a current project. The crowd was largely (unsurprisingly) PH (public health) and vendors selling their AI wares.

What I heard a lot of is “people trust a human in the loop”.

There was no evidence to back this up, and nor was there evidence that one human is as good as another.

Projects in flight didn’t differentiate either. I quizzed vendors and consultants on their processes to engender trust. They just copied and pasted from existing ethics and governance. The belief was that trust transfers over to technologically advanced systems. Errors in current processes weren’t fixed; new problems weren’t addressed; opportunities for improvement were considered incremental at best. Very little was evidence-based learning. nothing I saw was built on current literature in psychology, anthropology, consumer behavior, or health promotion. Or, oddly, public health.

So, I am starting a movement.

We are going to hear “AI with human in the loop” A LOT. Mainly because no one trusts AI unless they “let them in” to their process. The humans using AI trust the AI they have chosen to us, not the AI thrust upon them. Anyone who knows anything about consumer behaviour and customer experience would say “no surprises there”.

So, every time someone says “Human in the loop…” in reference to AI, respond with…

WHICH HUMAN.

It’s a simple questions, and one the hucksters and vendors, and influencers hadn’t thought of.