Keeping humans “in the loop” is often just a fig leaf for organizations too scared to let AI do its job. Look, I get it—nobody wants to feel obsolete. But the truth is, most “human oversight” is a joke. Take those self-driving cars that require a human to grab the wheel in an emergency. Has anyone actually done that in time? No. It’s a placebo for regulators and a distraction for engineers who should be focusing on making AI better, not pandering to our egos.

Here’s the thing: humans are great at creativity, empathy, and moral judgment. We’re terrible at processing vast amounts of data quickly and accurately. If you’re in a situation where a machine can make a better decision than you, let it. The “human in the loop” concept is often just theatre, a way to make us feel in control when we’re really just slowing things down.

But there’s a flip side. When it comes to ethical dilemmas or situations requiring nuanced understanding, humans are indispensable. AI can’t yet grasp the subtleties of human experience or cultural context. That’s where keeping humans in the loop makes sense—not as a safety net, but as a crucial part of the decision-making process.

So, here’s the deal: if you’re using humans to rubber-stamp AI decisions, cut it out. But if you’re leveraging human insight where it truly matters, keep it up. Otherwise, you’re just putting on a show.

Humans in the loop? More like humans in the way.