---
id: ins_kesava-mandiga-workflow-collapse-not-speedup
title: 'AI-native isn''t doing the same job faster, it''s collapsing roles'
operator: Kesava Mandiga
operator_role: 'Head of PMM, JustCall (SaaS Labs)'
source_url: https://www.linkedin.com/in/k3sava/
source_type: thread
source_title: 'Kesava on workflow collapse vs workflow speedup'
source_date: 2026-04-29
captured_date: 2026-05-03
domain: [ai-native, pmm, gtm]
lifecycle: [org-design, function-rebuild]
maturity: frontier
artifact_class: framework
score: { originality: 4, specificity: 4, evidence: 3, transferability: 5, source: 4 }
tier: A
related: [ins_kesava-mandiga-substrate-driven-intelligence, ins_kesava-mandiga-cheap-tier-delegation]
raw_ref: ../wiki/concepts/ai-native-pmm-practice.md
---

# AI-native isn't doing the same job faster, it's collapsing roles

## Claim

People keep using AI to do the existing job 2x faster. That's not the win. The win is that the function rebuilds around the substrate, PMM and design and content and competitive aren't separate roles anymore, they're loops on the same data layer that one person plus a model can run. The org chart goes from 7 PMMs to 2 super-ICs and a flywheel. The metric that matters is revenue per employee, not headcount.

## Mechanism

Speedup compounds linearly. Collapse compounds quadratically because every adjacent role you absorb removes a coordination tax. The first AI-native PMM at a company displaces three layers of process. The seventh one fights for budget.