About this episode: AI has made it possible for any one of us to make or draft almost anything on our own, but the cost is that teams have never been more isolated. Jakob Knutzen is Head of Product for Collaboration at Miro, and believes the best work now happens after everyone has gone off with their own LLM, when a team slows down, comes back together, and decides with their own thinking what they ought to do.
Jakob leads Miro’s professional services vertical, focused on making Miro great for agencies and consulting companies like Crema. In this episode, he and George Brooks talk about what AI has made easier and harder. Chatting on everything from helping teams get back to ‘multiplayer mode' to why UX craft and product sense are that much more important.
Everyone became an island again
George felt it inside Crema this time last year (in 2025). Our whole value proposition is cross-functional teams, it’s all about getting in a room and ideating and coming to an aligned decision. Then everyone started going to their LLM of choice and jamming in isolation.
“AI is incredible. Like we can do so much with it, but it’s also the most isolating moment that I think I’ve seen... everyone just became an island.”
George even tried to build a little tool called Co AI to bring everyone’s work into a shared canvas, and he’ll readily admit “it was terrible and never got it to work,” followed by the very relatable thought, “dang it, I wish I could do this in Miro.” So it’s been gratifying to watch Miro come at the same problem with its MCP, which Jakob says “has just proved to be incredible” because it’s “a way that you get from single player to multiplayer.”
He uses it in his own workshops, writing the structure and outcomes himself first, refining them into a script with Miro’s Sidekick or Claude, and then inviting his team to hook up their own Claude to the board and chuck in their individual artifacts before the conversation starts.
Generic is easy, but great UX has always been about less
The conversation turned to the design craft, which George sees as one of the last disciplines to jump into AI, with designers who still “really love moving pixels” and want to use their tradecraft as long as possible. That’s understandable when so much of good UX is about making something usable, functional, lovable, and slightly different from the competitor, and as George puts it, “AI is great at the common denominator. It’s great at generic.” At the same time, a lot of organizations are laying off their UX departments, and Crema is seeing an influx of demand for UX services as a result.
“They don’t realize that they’re just building generic. It’s not actually going to solve a unique problem in the world.”
When there’s no constraint, everybody builds everything that they want. Everything and more than they should. Jakob agreed that great UX has always been about less is more. And that product sense is critical now. Without it, teams end up building monstrosities that “just like have tentacles into everything, with a gazillion menus,” and they think they’re amazing.
Meeting people where they already are
Crema has used Miro since it was called RealtimeBoard, and early on we built a little app so executives who didn’t want to learn the tool could text in their ideas and watch them land on the board as sticky notes. Jakob spent his first year at Miro solving the same problem with Miro Engage, which lets people scan a QR code and participate, and which carries a lot of Butter’s legacy inside Miro. He knows that “so many people still find navigating like a two and a half dimensional whiteboard incredibly difficult.”
“You don’t want the tool to become the blocker, right? You want them to be the accelerator.”
That instinct shapes how his team thinks about the product today. George was candid about watching Miro grow up, remembering how “it just worked” in the early days before there was pressure to add this thing then that, and another. It’s a pattern he sees in Figma too, even as there is pressure to grow into new markets. Jakob sees most SaaS products moving in waves, expanding and then tightening, as they work through the straddle of “optimizing for your current user base versus searching for a new user base.”
Not all friction’s bad
Jakob is the first to admit that “as product people we often think of a product feature or functionality as the solution to all the problems in the world.” Even so, his team is thinking about conscious friction, small moments in the product that get people to slow down and think. For example, given the amount of AI slop he sees in Slack, a prompt asking “are you sure this is not an AI post?” is friction Jakob would really love to see. He personally keeps this rule for himself.
“AI should never draft anything that you post. You should always be the one that drafts. It’s fine if it edits.”
He’s also honest about the tension in that. AI has made him dramatically more productive, and he treats Granola, Slack, Miro, and Google together as a second brain he no longer has to spend time retaining, to the point that he gets “nervous if I have a call where I don’t have a Granola recorded.”
George named a weightier question underneath it, that “we’re kind of reprogramming our brains to go, you don’t have to remember this,” but Jakob pointed out that Socrates worried writing would ruin our memories too. So maybe it’s about balance! After a ten-week sabbatical (without touching AI) George came back to writing his lists in a notebook which slows him down enough to ask what really matters.
People of Product is brought to you by Crema — a design & technology consultancy and official Miro Solution Partner