How to Get a ‘Transformative’ AI Fluency Rating as a PM, with Wade Foster | CEO of Zapier

Aakash Gupta (0:00)

Every PM is about to get graded on AI fluency. Zapier, Shopify, and many others already are.

Wade Foster (0:06)

If AI was important in 2026, it's just the standard expectation for every job.

Aakash Gupta (0:10)

Wade Foster is the CEO and co-founder of Zapier. He wrote the viral AI fluency rubric that arguably kicked off this trend.

Wade Foster (0:18)

There is still something to be said about really understanding the customer problem, really having some intuition around how to go solve those problems. And if you combine those, you get excellence. But if you just let AI kind of just do everything, you're probably going to get mediocrity. You're going to get what people call AI slop.

Aakash Gupta (0:34)

What was capable last year that's now unacceptable? What does a PM that's transformative actually do differently on a Tuesday?

Wade Foster (0:42)

The line we use at Zapier on slop, you can delegate the work, but you can't delegate the accountability.

Aakash Gupta (0:46)

Looking at air table, now it's less than 10% of that amount. For you at Zapier, you've been around since 2012 Where does Zapier's business sit now?

Wade Foster (0:55)

One thing that's unique about Zapier is the valuation thing doesn't matter as much to us.

Aakash Gupta (1:00)

If you saw me running a product team like this, how would you grade my output?

Wade Foster (1:04)

If I'm spotting things like are obvious to me and not to you, I'm like, that makes me nervous, and I'm like.

Aakash Gupta (1:19)

Before we get into today's show, please take a second to check that you're subscribed on YouTube and following on Apple and Spotify podcasts. If you want access to all of my favorite AI tools, I've gotten them to give you an entire year of their paid plans. Check out bundle.akashg.com for an entire year of bolt.new, Airtable, Speechify, Descript, Magic Patterns, Linear, Dovetail, Arize, and Mobin. Now, into today's show. I've been trying to figure out what AI fluency actually means for PMs inside of companies. How do product leaders, how do CEOs measure you? And I could think of no better person than Wade. Wade published what I think was the first AI fluency rubric online. Many companies modeled it. Since then, Zapier has updated their thinking. How did they update their thinking? What actually matters? Day to day, how is Wade actually grading employees? In today's episode, he's not just gonna walk through his rubric theoretically. We'll actually show you some AI behaviors, and Wade will give his take on is this transformative? Is this unacceptable? He'll also talk to you about how these expectations have changed over time, so that you can meet the bar at your company or exceed it. Wade, thanks for being on the podcast.

Wade Foster (2:37)

Yeah, thanks for having me.

Aakash Gupta (2:39)

So, I have been dying to do this recording. I think we've rescheduled a few times, because I think you are the world's best person to talk to us about AI fluency. How do you guys measure AI fluency at Zapier?

Wade Foster (2:52)

Sure, might be most helpful to actually simply just show. So we have an AI fluency rubric. This, published March 31st, 2026, is our updated AI fluency bar at Zapier. So this is the V2. And you can see here, we basically break things down into four categories. You actually don't see, you see three on the screen here, capable, adoptive and transformative. We also have an unacceptable one, which is just simply the bar that we choose not to hire at. Obviously, for any company, you can tweak and tune these to fit your needs. We felt it was really helpful for us to really name some of the skills and behaviors that help people stand out in terms of their AI fluency. Maybe this is a product podcast, so let's maybe zoom down here to the product one.

I'll be the first to share, this is the V2 that we've shared externally, but we've even updated some of these things internally since. This space moves fast and you do have to constantly be tuning and tweaking what you're doing. You can see here some of the characteristics that make someone unacceptable, capable, adoptive, and transformative at this work.

In 2026, if you're using AI to do simple tasks, summarization, writing, looking up information, generating basic PRDs, things like that, that's kind of table stakes, I think, for most companies at this point in time. We're usually looking for somebody that's operating at a meaningfully higher level. So we want folks that have a really structured approach for generating specs, prototypes, things that they're going to reuse and refine across projects.