The Secret to AI Adoption Isn’t Better Training

Mandating AI adoption and measuring it in tokens burned is not a strategy. It's how you get employees faking usage to hit KPIs. Rhythmic CEO Cris Daniluk explains why AI adoption inside companies is messier and more interesting than most leaders expect. It won’t break cleanly along generational lines. Training won’t fix it. The people who naturally take to AI aren’t necessarily the youngest people in the room, and the people who resist it aren’t necessarily the oldest. AI is analog in a way that decades of digital transformation thinking didn’t prepare most organizations for. [2:15]

Key Takeaways

  • AI is analog, not digital. After years of “digital transformation,” leaders have to unlearn the idea that everything with a computer has to be expressed in binary commands.
  • Adoption will be uneven, and not along age lines. Some people instinctively get it, some never will, and most are in between, waiting for the moment they see it’s more than a better Google.
  • You can’t train or force your way to adoption. Cramming AI down people’s throats doesn’t produce a good result.
  • Don’t incentivize token usage. Some employees are faking AI usage to hit token KPIs, which is like rewarding a steel plant for burning electricity. It also keeps people from growing into better patterns.
  • Give early adopters grace on efficiency, and find champions in different parts of the business to show everyone else the way.

Cris (00:00)
I think it’s important for people to realize that, especially leaders, that we’ve been forced into this digital mindset for so long. We had phrases like digital transformation, and we had this notion that you have to express everything in a binary world and AI doesn’t work that way. AI is analog. AI has much more depth to it.

Some people will naturally gravitate toward that. It makes sense to them to talk to a computer. It never made sense to them to have to adopt language and coding and commands and things like that to interact with a machine. Other people are so conditioned by the digital world. And this is where I think Gen X has a little bit of an unfair advantage in AI. We grew up with the foot in both worlds. We had the opportunity to leverage this digital world, but we didn’t have to be defined by it.

And so people are going to have an uneven adoption curve. It’s not going to be like your younger employees are going to just dive on the AI train and your older employees are going to resist it. It’s much more complicated than that. And you’re not going to solve it with training. There will be people that instinctively get it. There will be people that never get it. And most people will be in between where they need to have the moment where they see that ⁓ this isn’t just a toy.

I can actually do something meaningful in my life. It’s not just a better version of Google. So you can’t really cram it down people’s throats and expect a good result. There are actually people that are using tools to just fake burn AI tokens so that they can hit their token budgets that they need to for their personal KPIs. That’s bananas.

And incentivizing that is like incentivizing a steel plant to use as much electricity as possible. They need as much output as possible. And electricity is a byproduct of that, but they shouldn’t be using more electricity than they need to because it’s already an enormous amount. And so you want to give users adopters some grace in their token usage.

That they’re going to be inefficient and ineffective early on and use more tokens than the next person who’s a little bit further down the curve to accomplish the same task. But at the same time, incentivizing tokens as an output means that you don’t have those users then growing into better patterns and ways to work. I think a much better approach is to find champions inside different parts of the business that can be those early adopters and show people the way.

Scroll to Top