IISc DoMS / March 6, 2026

AI & The New
Management Playbook

Ashish Muralidharan Product Manager, Microsoft
A question for the room

"How many of you feel anxious about your future, your career, your relevance, because of AI?"

Raise your hands
Act I
THE TENSION
The world we operate in

VUCA Framework

Volatility
Rapid, unpredictable change
Uncertainty
Unknown outcomes and paths
Complexity
Interconnected, layered systems
Ambiguity
Multiple interpretations of reality

"It's a one-way street going upwards."

Historical pattern

The Pattern of Disruption

Shipping
15th century
Industrial Revolution
18th century
💻
Information Technology
20th century
🤖
Artificial Intelligence
Now
Factors of production

The Economic Equation

Land
Finite resource
Labor
Being replaced
Capital
Still essential
Technology
The new factor

"Technology is fundamentally replacing labor."

Act II
THE PATTERN
The cycle repeats

Technology vs. Labor

1
Technology Arrives
A new capability emerges
2
Physical Displaced
Manual labor automated
3
Operational Displaced
Process work eliminated
4
Intellectual Displaced
Cognitive work at risk
Displacement by category

Running Out of Rungs

Physical Labor 95%
Operational Labor 72%
Intellectual Labor 45%
??? 15%

"If you don't need intellectual labor anymore, what's left?"

The final frontier

Robots & The Physical Frontier

🤖
Tesla Optimus
Humanoid general purpose
🚶
Unitree
Agile bipedal robots
Hyundai / Boston
Industrial robotics

"Complete replacement across every dimension."

OpenAI classification

5 Levels of AGI

Level Capability Description Status
L1 Conversational AI Natural language interaction Done
L2 Reasoning AI Complex problem solving Done
L3 Agents Autonomous task execution HERE
L4 AI Organizations Multi-agent coordination Emerging
L5 AGI General intelligence Unknown
Definition

What Is An Agent?

"A system that can pursue a goal by reasoning, planning, using tools, and continuously learning."

Act III
GROUND ZERO
Software
The first domino

Why Software First?

100%
Digital
Entirely bits, no atoms
End-to-End
Signal-to-closure fully automated
📊
Rich Data
Decades of training signal
The shift

"Single engineers spin up 20 agents that build the entire product end-to-end."

Industry voices

Voices from the Field

Andrej Karpathy

"The hottest new programming language is English."

On "vibe coding" and the future of software

Naval Ravikant

"PM is vibe coding. Engineering is prompt engineering."

On the convergence of roles

The economics of creation
$0

"The cost to build software has come down to zero."

If building is free...

The New Battle

Building
Commoditized
Distribution & Marketing
The new competitive moat

"The new job for everyone has become marketing."

Already happening

Real Consequences

50%
Block Layoffs
Half the workforce cut
-30%
IT Stocks
Indian IT down 25-30%
Innovator's Dilemma
Playing out in real-time
Act IV
THE ECONOMICS
The optimistic case

The Abundance Thesis

🤖
Remove Humans
From production
📈
Deflation
Costs collapse
💵
Goods Cheaper
For everyone
🌎
Era of Abundance
Post-scarcity
Follow the money

The AI Value Chain

Energy
📈
Chips
Fabricators
🧠
Frontier Labs
Hyperscalers
📱
App Layer
🎯
Distribution
Not where you think

Where Value Lives

The Model
Commoditizing fast
Compute
Infrastructure advantage
Context
Data and domain knowledge
Distribution
Reaching users at scale

"Agents don't need UI."

Act V
FROM THE INSIDE
What I've seen

Microsoft's AI Journey

What Failed
Top-Down Mandates
Executive directives alone didn't move the needle
What Worked
Bottom-Up Adoption
Empowering individual contributors
Key enabler
Unlimited Compute
Key enabler
Incentive Alignment
Result
20x Productivity
Inside the machine

Project Nebula

100
Product Managers
600
Engineers
50
Designers
Weekly AI Upskilling
"Tribal knowledge." The best practices are discovered, shared, and iterated on weekly.
The key insight

"The most valuable managers will be the ones who figure out how to diffuse AI throughout the company."

Act VI
THE NEW PLAYBOOK
Organizational design

AI Organizations

Design workflows for agents like managers designed them for humans.

BEFORE
PM writes spec
Manual doc, 2-3 days
NOW
PM prompts agent
Agent drafts spec, 30 min
NEXT
Agent orchestrates agents
PM reviews output, 10 min
What matters now

Two Skills That Matter

01
Prompting
Instruct precisely. Decompose problems. Provide context. Iterate on output.
02
Evaluations
Measure AI quality. Build rubrics. Assess hallucinations. Know when to trust.
The new archetype

The Age of the Polymath

"Everyone is becoming a generalist."

The one-person billion-dollar startup is no longer a thought experiment.

Your action plan

What You Should Do

01
AI Budget
Dedicate time and money to AI tools. Treat it as essential infrastructure, not a nice-to-have.
02
Build AI Fluency
Use AI daily. Prompt, iterate, evaluate. Build muscle memory for working with agents.
03
Think in Workflows
Map your processes. Find where AI agents can slot in. Design for delegation.
04
Become a Generalist
Learn everything. The polymath wins. AI handles depth; you provide breadth and judgment.

Remember that anxiety you felt?

Go build something.

Ashish Muralidharan

Product Manager, Microsoft

1 / 34