Working Surfacelink · 20 Nov 2026
01
A factory press stamps out a long line of identical moulded chairs onto a conveyor, while one hand-carved chair with an unusual shape stands apart on a pedestal and a figure seen from behind leans in to study it with a magnifying glass, the whole scene drawn with wide empty margins on every side.
Thursday Thoughts on AI

What I Look For in an AI PM - Part 4: The Prototype Isn’t Enough Anymore

Jaclyn Konzelmann · 19 November 2026

Company story · prototyping · roles · After the prototype

Read the original
Takeaway

Jaclyn Konzelmann, a product director at Google Labs, no longer counts a working prototype as proof when hiring product managers, because AI makes generic ones easy to build.

Summary

Jaclyn Konzelmann leads product work at Google Labs, the company's experimental product group, and hires AI product managers. Earlier in 2026 she advised candidates to bring a working prototype to interviews. In a post published on 19 November she says that is no longer enough, because AI tools produce similar, generic prototypes from similar requests.

Key points
  • She gave two similar requests for a fitness web app to three AI tools, and the six results looked much alike.
  • She judges generic output by excess settings and buttons that a real user would not need.
  • In a demonstration she asks who the product is for, what was left out, whether the core loop works and why a user would return in week three.
  • She describes her own children's maths game, where watching her daughter tap at random led her to freeze the board for 1.5 seconds after a wrong answer.
Implication

A product team can review its own prototypes with the same questions about audience, omissions and the core loop. Using them beyond hiring is an inference and is not the author's claim.

Plays
Product

When reviewing a prototype, ask its maker who it is for, what they left out and why a user would come back in week three.

Derived by Working Surface from the article; more in the Playbook. Source line: The prototype used to be the proof.

Source issue

20 November 2026: as AI agents make prototypes, design tools and code cheap, the scarce work becomes a person's judgement and a named person answering for the result.