Human in the Loop: What the Data Says About AI and ROI

Not long after ChatGPT fell on us back in 2023, my colleagues and I started scoping AI training courses. What would people need to understand? How could they gain competence and competitive advantage in a few hours of on-demand education? As it turned out, we were a bit early to the game. Leadership at client companies didn’t know what they didn’t know, and, being honest, didn’t want to train people on a technology they hadn’t vetted (and didn’t understand themselves).

In the mother of all understatements, a fair bit has happened since then, and yet many companies are still looking for the right kinds of training to unlock the human side of their cyborg teams.

Max Haining has answers. Max is the founder of 100 School, an AI capability business. He’s spent the last two years studying what separates the teams that have genuinely changed how they work from the ones still waiting for the tools to do it for them. The gap, he argues, is behaviour.

Most organisations keep responding to the problem by adding more: more workshops, more features, more content. Max argues this compounds the chaos rather than cutting through it. What the best teams have done is almost the opposite: gone two steps back from the features entirely, and focused on building the thinking habits that let people adapt independently. You can read the full interview below.

Our research data this week is about people too. ‘Humans in the loop’ aren’t a sign of weak AI workflows. On the contrary, we find that the companies that are successfully extracting real ROI tend to have more human intervention in their core AI workflows than their peers, not less.

Both threads are on the table next Tuesday at #growthLIVE, which is built around exactly this problem: how revenue-focused organisations move from AI deployment to AI results. If you’re coming, we’ll see you there. If you haven’t applied to attend yet, don’t delay.

aibl Research: The Human Conditions for AI-Driven Efficiency

Our data drop this week is a teaser for #growthLIVE next Tuesday, and our message is dead simple: the definition of AI success isn’t to get people out of the loop.

The data back up what we’ve been hearing in our interviews with successful operators: The companies focused on revenue growth that report positive ROI actually have more human intervention in their core AI workflows than their peers, not less.

We’re still at the stage where people have more institutional knowledge, instinct, and creativity than any AI system. The point isn’t to automate what people can still do better…it’s to restructure how time is spent and value is created.