An improv class I still think about
What an improv class taught me about adapting to AI. Plus an "almost possible" list to try.
By Mollie Amkraut Mueller · · 5 min read
Learning AI can feel like a slightly unfair assignment. You get comfortable with a tool, figure out what it's good for, maybe even teach someone else how to use it. Then something changes and you're figuring it out again.
I hear this from leaders all the time. It's a big topic in the advising work I do with executive teams and people leaders: when the tools and capabilities keep changing, which skills will have longevity?
The capability I keep coming back to is agility. Being able to encounter something unfamiliar, tolerate the uncomfortable bit, and figure out what you can do with it.
Which brings me to an improv class in London.
Back in 2017, I was working on an IDEO project about helping small businesses adapt to the accelerating pace of change. How could they keep up with the big guys, and ideally thrive?
(Yes, 2017. We were very concerned about how fast everything was moving. Nine years later, that pace of change sounds like a lovely break. We had no idea what was coming.)
To get ourselves and the client team into the right headspace, we tried a bunch of different inspiration activities. One was an hour of improv. The acting kind.
I find improv pretty uncomfortable. You're on the spot. People are watching. Someone introduces a completely random plot twist, and you have to respond with whatever you've got.
Your instinct might be to freeze, reject the premise, or wish you'd chosen a different team activity.
But the exercise asks you to accept what's been introduced and build on it. Add something. See where it goes, even though you have very little idea where that might be.
We were practicing the feeling of responding to change.
I think about that class all the time now.
With AI, we're asking people to learn new things while the things themselves keep changing. That requires a willingness to be a beginner repeatedly, including at work, where we're generally quite attached to being competent.
It's easy to say "have a growth mindset." It's harder when you've just spent an afternoon (or your entire career) figuring something out and someone tells you there's already a better way.
I still think improv would be useful practice for a lot of us. But there's also something you can try that doesn't involve performing in front of your colleagues.
Keep an "almost possible" list
I heard this idea from Josh Woodward, who leads Google Labs, on Lenny's Podcast. His team keeps a running list of things technology can almost do. They watch for the moment something becomes possible, so they're ready to experiment with what it could unlock.
I've started doing the same, and it's helped me think through my strategy and what to plan for next. It's also been surprisingly useful for digesting change. Keeping the list gets me thinking about what might be coming and what it could mean for my work before it arrives.
You could start with something you've tried to do with AI that didn't quite work. Maybe it needed too much supervision, or the output wasn't consistently good enough. Put it on your list, along with what would need to improve for it to become useful.
Then take it a step further: if that became possible, what would change for your work, your organization, or your market? What would you want to try?
The next time a new capability comes along, you've already given yourself a head start on thinking about it. And potentially a reason to be excited.