ProductTank AMS: How Miro's Ben Stein shifted to prototype-first PM with AI

Ben Stein is a Senior Product Manager at Miro, the visual collaboration platform. In a session published by ProductTank AMS in May 2026, he described a shift in how he approaches product work — moving away from traditional specification-first thinking and toward a prototype-first approach enabled almost entirely by AI tools.

The context matters. Miro builds tools used by product teams, so Stein works in an environment that is both AI-native and unusually close to product methodology conversations. What he describes is not a pilot or an experiment — it is how he actually runs day-to-day product work in 2026.

From specs to working prototypes

The centerpiece of the shift is using AI to move directly from a rough concept to a functional, interactive prototype without writing code. Stein built the Miro Engage prototype — a feature for facilitating structured collaboration — almost entirely through AI-assisted workflows. The resulting prototype was usable enough for stakeholder feedback without design polish or engineering involvement.

The practical consequence of this approach is that the question “what should we build?” becomes answerable with less upfront investment. Because building costs less, teams can test more hypotheses before committing. Stein’s argument is that this changes which PM skills matter most: the ability to reject ideas quickly and identify which prototypes reveal real signal becomes more valuable than the ability to write thorough specifications.

Automating the repetitive layer

Alongside prototyping, Stein built a weekly update workflow by connecting Jira, GitHub, Slack, and meeting transcripts. The workflow pulls activity across those sources and prepares a ready-to-share summary. A task that previously took 30 to 60 minutes of manual aggregation now takes a few minutes of review and editing.

This is representative of a pattern that appears across multiple 2026 case studies: the highest-value AI automation targets the coordination and synthesis work that occurs between meaningful decisions. Writing a summary of what happened is not judgment — it is aggregation. Removing it from the PM’s plate does not reduce PM value; it redirects attention toward where judgment is actually needed.

What the shift requires

The article is candid about what changes when prototyping and synthesis become faster. The premium on problem selection increases. Choosing which hypothesis to test, which prototype to show stakeholders, which feedback to act on — these decisions carry more weight when acting on them costs less. Stein’s framing: “When building becomes easier, the real superpower is deciding what not to build.”

Teams that benefit most from this pattern tend to have clear product principles, enough direct user access to evaluate prototypes quickly, and a willingness to discard work before it becomes a commitment. Teams without those conditions may build more things faster without necessarily building better things.

Who this is useful for

Product managers working at mid-size B2B or SaaS companies, particularly those involved in internal tooling or collaboration software, will find Stein’s account directly applicable. Managers at larger companies where prototype fidelity expectations are higher may need to adjust the approach, but the underlying workflow automation principles transfer broadly.