Laura Burkhauser on helping teams rewire for AI

When Laura Burkhauser talks about AI adoption, she doesn’t sound like a consultant. She sounds like someone who’s been in the trenches. “We talk about a tech acceptance scale at Descript,” she says. “It goes from hostile to skeptical to converted to rewired.”

The metaphor fits Descript, where Laura leads the product team, helping make audio and video editing as easy as working in a doc. But even there, not everyone jumped in at once. “Some people were ready to strap on a Neuralink,” she says, “and others were proudly old-school.”

That mix of eager and hesitant is exactly what she helps teams navigate.

1. The real AI adoption curve: hostile → skeptical → converted → rewired

Laura’s framework sounds simple, but it resonates because it mirrors what she sees every day. She describes four mindsets people move through as new technology lands:

Hostile is the reflex stage — the “This isn’t for me” reaction. It’s the designer who quietly refuses to open a new tool, or the engineer who jokes that AI will just make things worse. Underneath it is fear: a sense that adopting it means erasing part of your craft.

Skeptical comes next. These describe people who’ve lived through too many hype cycles to trust the noise. They’re curious, but cautious. They’ll say, “Show me it actually helps, not just that it exists.” These are often the voices of experience, and Laura believes they’re crucial guardrails against blind adoption.

Converted is where most teams are right now. They’ve accepted that AI is real and here to stay, but their day-to-day hasn’t changed much. They use the tools, but around the edges; a little prompting here, a draft there. “They believe it,” she says, “but they’re not living it.”

And then there’s Rewired: the small but growing group who’ve let the technology reshape their habits. They no longer think, “How can I use AI?” They just use it, seamlessly, the way you open Slack or Google Docs without thinking.

For Laura, the work of leadership is to move teams along that curve, not by preaching the future, but by giving people experiences that make it felt. “You can’t convince someone into being rewired,” she says. “You have to show them.”

2. The ‘burning platform’ vs. the beach

Change management theory talks about two motivators: the “burning platform” (what pushes you to move) and the “beach” (what pulls you toward something better).

Laura says AI adoption has both, but too many teams focus on the fear side. “The FOMO is real,” she says. “Everyone’s repeating, ‘I have to adapt or be left behind.’ But no one does smart things when they’re motivated by fear.”

Her advice is to find your beach. “Find the thing that makes you excited about this stuff. That’s what rewires you. Not panic, excitement.”

For Laura, that “beach moment” came when she used Cursor, a vibe-coding tool connected to her team’s codebase. “I built an undo button that shipped to production in one shot,” she says. “It was like getting a superpower.”

3. The Matrix moment: helping your team ‘know kung fu’

Once Laura had her own lightbulb moment, she wanted her team to feel it too. “I didn’t want AI to be something people read about in Slack threads,” she says. “I wanted them to touch it.”

So she gathered her product managers and designers and gave them a simple challenge: pick something small you’ve always wanted to build — an experiment you’ve never had time or resources for — and try to prototype it using AI. No pitch decks, no approvals. Just curiosity and a few hours to play.

At first, there was hesitation. Some asked if this was really what “using AI” looked like for product managers. “But as soon as a designer can QA their own feature or a PM can prototype something they’ve been waiting months to resource, they’re in,” she says. “It’s like that scene in The Matrix — you plug in and suddenly you know kung fu.”

Even if none of the projects shipped, something more important happened: confidence spread. The team had gone from theoretical believers to hands-on practitioners. “That’s the difference between being converted and being rewired,” Laura says. “It’s not about drinking the Kool-Aid. It’s about giving people a firsthand taste of what’s possible with AI, and letting that excitement pull them forward.”

4. Turning excitement into structure

Of course, enthusiasm isn’t strategy. Once the team had their “beach” moment — the sense of excitement and possibility — Laura wanted to channel it into something structured and measurable.

So she gathered her product, design, and engineering leads for a reflective exercise. The prompt was simple:

“I wish I could spend less time doing ___ so I could spend more time doing ___.”

It was the kind of question that doesn’t require a workshop or a whiteboard, just honesty. And the answers, Laura says, were remarkably consistent.

• On the “less” side: progress updates, changelogs, documentation, post-launch reports, and all the routine administrative work that quietly fills a PM’s week
• On the “more” side: talking to customers, thinking strategically, shaping long-term bets

What struck Laura most was how universal the pattern was. “When you look at what people want to spend more time on,” she said, “it’s all stuff that’s self-evidently better for our customers and for Descript.”

From there, the exercise became a design problem. Together, the team mapped every “less” item to what could be automated or accelerated with AI: the areas where automation would create immediate relief. Then they mapped every “more” item to what AI could amplify: the parts of the craft that could get richer, not just faster.

Those two lists became the backbone of Descript’s internal AI roadmap. It was a practical plan for where automation could free up time, and where human creativity could push the work forward.

“That’s how we set our goal to double EPD (Engineering, Product, and Design) output by the end of the year,” says Laura. “Not by squeezing people, but by removing the drag so they could do more of the work that actually matters.”

5. Simulate → automate → delegate

To turn those insights into practice, Laura creates a clear three-stage path for her team: Simulate → Automate → Delegate.

Simulate comes first. Teams run small, manual experiments to see if a workflow is worth automating. “Don’t start by buying tools,” Laura says. “Start by simulating the workflow you want.” In practice, her PMs and designers use Cursor or Claude to test their own ideas, like drafting changelogs, updating PRDs, or writing post-launch reports. The goal is to learn, not to be perfect.

When a workflow consistently saves time or improves quality, it moves to the Automate stage. Here, the team decides whether to build it in-house or buy something off the shelf. “We’re PMs; we know this,” Laura says. “You start with an MVP. You verify the workflow is a good one before you invest.”

Finally comes Delegate. This is when AI handles full workflows with a human still in the loop. “Over time, we want to trust the system to take on more of the work,” Laura explains, “but never without oversight.”

This loop of simulation, automation, and delegation anchors how Descript experiments. Each week, Laura’s team reviews what worked, what failed, and what’s ready for the next stage. On Fridays, they host a casual vibe-coding show and tell where PMs, designers, and engineers share their latest experiments.

“It’s not about showing off,” Laura says. “It’s about normalizing experimentation and turning curiosity into a habit.”

The framework gives the team a shared language for progress. Even a scrappy manual test counts as forward motion. As the experiments add up, so does confidence. And fluency follows, one workflow at a time.

6. Responsible AI starts with values

Descript’s tools make powerful things possible. A user can edit a podcast just by changing the transcript, or swap a single word in their audio and have the speaker’s lips automatically update to match. It feels like magic — the kind of leap that can make anyone feel like a professional creator.

But Laura is quick to point out that power cuts both ways. “You can imagine the abuse vectors,” she says. “We want to make sure if your use case is a bad-actor use case, we’re not the platform you choose.”

That tension between creativity and control is something Descript’s team talks about constantly. They know their tools can amplify good work, but they can also be misused for deception or harm. “When you make it easy to edit anything, you have to think about what people should do, not just what they can do,” Laura explains.

To stay grounded, the company uses a simple north star: Make creativity easier for good actors, and harder for bad ones.

That principle shapes everything from feature design to onboarding flows. Deepfake prevention isn’t an afterthought; it’s baked into product choices. The team invests in visible watermarks, content authenticity signals, and user education that reminds creators what ethical use looks like. “We want people to feel empowered,” Laura says, “but we also want them to understand the responsibility that comes with that empowerment.”

It’s the same philosophy that guides how Descript uses AI internally: scale what’s human, not what’s hollow. The tools might speed up the work, but trust, taste, and intent still have to come from people.

“AI can help you sound like yourself on your best day,” says Laura. “But it can’t decide who you are.”

7. Advice for new PMs in an AI-native world

Laura’s advice to people entering the product world now is simple: Earn your way into the room where context is shared.

“Context is currency,” she says. “That was true before AI, and it’s even more true now.”

When she started her career, value came from making slides and taking notes — the work that got you into the meeting where the real context lived. “Now, AI can do that work. So your job is to bring something else: curiosity, creativity, the ability to connect dots.”

Her bottom line is that the skills will change, but the craft stays the same. “If you love building, if you care about customers, and if you can stay curious, there’s a place for you. The roles will evolve, but the heart of product work won’t.”

Keep learning

Laura’s framework appears alongside those of Kene Anoliefo, Ravi Mehta, Aakash Gupta, and Elena Verna in Atlassian’s new ebook, AI fluency: The new product superpower. Read it now →And you can listen to her full interview with Atlassian here.