Key Takeaways

  • Mike Krieger tested running projects without a PM at Anthropic, assuming Claude could handle technical coordination, but discovered critical handoffs immediately started dropping.
  • Ami Vora defines product management as a durable bridge between messy human problems and technological capabilities, requiring human judgment under incomplete data.
  • When engineering velocity increases through AI, operational demands rise because faster builds create more surface area for misaligned requirements.
  • Krieger outlines the convenor archetype: human leaders who coordinate both AI models and cross-functional teams toward a single outcome.

The Fallacy of the Self-Driving Project

When Mike Krieger moved into an individual contributor role at Anthropic, he wondered whether product managers were still necessary in an environment powered by Claude. Engineering tasks moved quickly, code generated on demand, and AI models handled complex synthesis.

Krieger questioned whether to staff a PM on his team: "I was like, 'I don't know. Like, do we need a PM? Like, Claude's got it. Like, we've got a lot going on.' And she's like, 'No, we really need a PM.' And it like such a reminder like once they joined like all of the things and all the glue and all of the connective tissue and all of the things that were about to get dropped if I had not done that would that would have happened."

Without a dedicated person holding the operational connective tissue together, projects crack at the seams. AI models can draft specs, generate roadmaps, and write unit tests, but they do not notice when legal, design, and go-to-market teams operate on conflicting assumptions.

The Convenor Archetype

Krieger points out that AI models cannot run human coordination. In a recent talk on human archetypes in an advanced AI era, he identified a central archetype: the convenor.

“I don't think claude is a convenor yet,” Krieger noted. “Somebody still needs to bring the clouds and the people together to like get the work done.”

Ami Vora echoed this perspective. The expansion of raw technical capability does not eliminate the need for taste and problem selection. As Vora put it, “I think the job of product is always to be a bridge between the real problems that people have in the world and the technology that you can use to solve it. And like that's always been true.”

When building software becomes cheaper and faster, the number of potential features explodes. Knowing what not to build becomes harder. Vora explained: “In some ways that just like expands the universe of what is possible to build and you still need to have really good judgment about like what to build based on sometimes pretty limited information cuz it feels like everything's moving really fast.”

Velocity Multiplies Operational Risk

Faster execution does not erase organizational friction; it amplifies it. If an engineering team ships three times faster with AI assistance, they run into three times as many unaligned stakeholders, edge cases, and distribution bottlenecks per week.

“I think all that's happened now is that because we can move faster, that role is like you have to be operationally excellent in that role in a way that like even beyond what you had to be before,” Krieger said. “But it definitely has not gone away.”

PMs do not protect their jobs by acting as ticket managers or status updaters. Claude does those tasks faster. PMs protect their value by owning context, resolving organizational ambiguity, and bringing humans and agents into alignment around user realities.

What to Do With This

Audit your product workflows this Friday. Identify every active project where an engineer or founder is acting as a de facto coordinator between automated tools, internal teams, and customer feedback. If project requirements or edge cases were dropped during your last sprint, stop expecting AI tools to catch them and assign a clear human convenor to own operational alignment.