AI Adoption Is a Culture, Not a Metric.

My 3-year-old nephew has a play kitchen. He is extremely confident in it. He will serve you a lunch of a stale Cheez-It and plastic tomato, garnished with a pen cap (the pen itself suspiciously missing) and look you dead in the eye like he just plated a Michelin star dish. The output is not edible. But, hey, I asked for lunch and he checked that box.

That's what AI usage mandates remind me of.

I'm talking about the companies that sound like: "Every employee must use AI tools X hours per week." It might seem forward-thinking, but it's your fastest path to workslop at scale.

You're not getting AI adoption. You're getting workslop.

When you tell people they have to use a tool a certain number of times, they'll paste things into ChatGPT that didn't need to be pasted into ChatGPT. They'll generate summaries no one reads. They'll auto-draft emails that sound so robotic it hurts to read. They'll hit their quota and move on.

Leaders think they're driving adoption. What they're actually getting is compliance. And not the good kind. Not the kind where people internalize something and it changes how they work. Nope. They get the kind where people hit a number and move on. Compliance without intent. And the output of that is slop.

The companies mandating AI metrics are the same ones who will be scratching their heads in eighteen months wondering why their internal docs all sound the same, why nobody can think through a problem without reaching for a prompt, and why the "innovation" they expected looks suspiciously like a bunch of slightly reworded first drafts.

They'll blame the tools. But the tools aren't the problem. The mandate is.

Transactions, not transformations

Here's what I'll concede: the mandate might push teams to move a little faster. Fine. But speed on the same work is not a competitive advantage. It's a marginal efficiency gain with a ceiling, and that ceiling is low.

The mandate frame treats AI like a faster conveyor belt and your people like the parts on it. Same job, same thinking, just more throughput. If that's how you see your team's value, you've already decided the human is the replaceable part. It's giving "my AI employee works 24/7 and never asks for a raise" LinkedIn energy. If that's your strategy, your people already know what you think of them.

I spent almost a decade in strategy consulting. I have the Deloitte alumni email to prove it. And our whole job was to make sure a client didn't just digitize a broken process and call it a digital transformation. But that's exactly what AI usage mandates are. Nobody's rethinking the work. The tool gets shinier. And someone in leadership calls it progress.

Hiring tech is the clearest example. The ATS itself makes sense. A central place to manage candidates, fine. But then keyword matching got bolted on. Automated rejections. Resume parsing that flattens humans into data fields. Technology that could reject a thousand people before lunch. And somehow that was ✨innovation✨. Recruiters and candidates have been screaming about it for a decade. The dashboards got prettier. The thinking didn't change.

That's the pattern AI usage mandates are repeating right now, across every function. You can ship features 30% faster than your competitors. Doesn't matter if your customer never wanted them.

What should happen instead: protected time to play

Here's what I'd actually respect seeing from a company: "You should spend X hours of your working week playing with a new AI tool of your choice. During normal hours, not on your lunch break, not at 9pm."

Not "use AI for your job." Play with it. Explore. Break things. Try the weird stuff that has nothing to do with your Q3 deliverables.

There's a massive difference between "use this tool to do your existing work faster" and "take time to understand what this technology actually is." The first produces workslop. The second produces people who can think.

That's how you build intuition. Not the vendor demo version of understanding a tool, but the real version. Where is it genuinely useful. Where is it confidently making things up. And what's possible that nobody's thought of yet, because nobody's had the unstructured time to stumble into it.

Play does something mandates fundamentally can't. It helps them see their work differently. When someone spends unstructured time with a tool that thinks in ways they don't, they start asking questions they wouldn't have asked before. Not "how do I do this faster" but "why are we doing this at all" and "what would this look like if we started from scratch." Now, that? That's a transformation.

The nonobvious dots

I'm a firm believer that you shouldn't just look in one domain to understand a topic. You should look everywhere else, too. And yes, I'm about to use a 2000s cheerleading movie to make a point about AI strategy. Stay with me.

In Bring It On, the Toros find out their former captain stole their entire routine from the East Compton Clovers. So they do the responsible thing. They hire a professional choreographer, Sparky Polastri, to build them something original. Spirit fingers and all. Here's the problem: Sparky sold the exact same routine to a bunch of other squads. So now they're back to square one, performing the same moves as everyone else.

Sound familiar?

That's what AI mandates produce. Everyone using the same tools the same way, generating the same outputs, and calling it a strategy. But instead of finding another shortcut, the Toros scrapped it. They went out and studied completely different dance styles: ballet, contemporary, stuff that had nothing to do with competitive cheer. And then they built something that was authentically theirs. They didn't study every winning routine and build a franken-routine out of the greatest hits. They found inspiration from completely different disciplines, challenged their own assumptions, and created something no one had seen before.

That is literally how we built Introve. The talent industry has its own Sparky Polastri. The same hiring playbook sold to every company, the same keyword-matching, the same behavioral interview scripts. Underwhelming and overplayed. So instead of optimizing around those margins, we started looking at every context we could find where someone is intentionally trying to connect a person to something they'll love. Realtors and how they match people to homes. Matchmakers and how they read people and discern if two people will work out. That cross-pollination gave us fundamentally different ideas. The kind that can't come from staring at your own industry's best practices.

This is exactly what happens when you let people play with AI instead of mandating they use it. Someone in operations plays with an image generator and starts thinking differently about how they communicate process changes. Someone in sales experiments with a reasoning model and realizes the way they've been qualifying leads is based on assumptions they've never tested. That's actually how we stumbled into building Mirra. We were playing with conversational AI and realized there's a massive gap between how candidates actually talk about their experience and how they've been taught to talk about it. That insight didn't come from a mandate. It came from unstructured time with a tool and a question we didn't know we were asking.

These are the nonobvious dots. And they only get connected by people who've been given the room to wander.

The real metric

If you need a metric, try this: measure how often your team surprises you with something you didn't expect. Yeah, that's a terrible metric. You can't put it in a dashboard. It's going to make some of you deeply uncomfortable. Good.

The companies that mandate AI usage will have beautiful adoption dashboards. The companies that protect time to play will have ideas nobody saw coming. In eighteen months, only one of those will matter.

Stop counting prompts. Start making room for play.