Ashish Muralidharan Product Manager, Microsoft IISc DoMS / March 6, 2026
"How many of you feel anxious about your future, your career, your relevance, because of AI?"
ACT I
THE TENSION
V
Volatility
Speed and magnitude of change are increasing exponentially
U
Uncertainty
Outcomes are becoming fundamentally unpredictable
C
Complexity
Interconnected systems create cascading effects
A
Ambiguity
The rules of the game are being rewritten in real time
"It's a one-way street going upwards."
SHIPPING 1500s INDUSTRIAL 1800s INFORMATION 1990s ARTIFICIAL INTELLIGENCE 2024+
Factor 1
Land
Factor 2
Labor
Factor 3
Capital
The Disruptor
Technology
"Technology is fundamentally replacing labor."
ACT II
THE PATTERN
Technology Arrives Physical Labor Displaced Operational Labor Displaced Intellectual Labor Displaced
??? 15% INTELLECTUAL 45% OPERATIONAL 72% PHYSICAL 95%
"If you don't need intellectual labor anymore, what's left?"
Tesla
Optimus
Humanoid robotics at scale
Unitree
G1 / H1
Affordable agile robotics
Hyundai
Atlas
Industrial-grade mobility
"Complete replacement across every dimension."
L5 Artificial General Intelligence Unknown timeline L4 AI Organizations Emerging signals L3 Agents We are HERE L2 Reasoning Achieved (o1, Claude, Gemini) DONE L1 Conversational AI Achieved (ChatGPT, Copilot) DONE
01 Signal Receives a goal 02 Cognition Reasons & plans 03 Action Uses tools 04 Closure Delivers & learns
"A system that can pursue a goal by reasoning, planning, using tools, and continuously learning."
ACT III
GROUND ZERO
Property 1
100% Digital
No atoms to move. Pure information, infinitely replicable.
Property 2
Signal to Closure
The entire workflow can be automated end-to-end by agents.
Property 3
Rich Training Data
Decades of code, docs, and patterns to learn from.
"Single engineers spin up 20 agents that build the entire product end-to-end."
Andrej Karpathy
"Vibe Coding"
You describe the intent. The AI writes the code. You just steer.
Naval Ravikant
"PM is Vibe Coding"
"Engineering is prompt engineering." The roles are converging.
"The cost to build software has come down to zero."
The Shift
Distribution & Marketing
When building costs nothing, the only moat is distribution. Getting to users is the game now.
"The new job for everyone has become marketing."
50%
Block Inc. layoffs across engineering
-30%
IT services stocks down in 12 months
Innovator's Dilemma
Playing out in real-time across industries
ACT IV
THE ECONOMICS
Step 1
Remove Humans
From Process
→
Step 2
Cost Deflation
→
Step 3
Goods Get
Cheaper
→
Outcome
Era of
Abundance
01 Energy 02 Chips 03 Fabricators 04 Frontier Labs 05 Hyperscalers 06 App Layer 07 Distribution
The Model
Models are commoditizing. This is not the moat.
Moat 1
Compute
Access to massive GPU infrastructure
Moat 2
Context
Proprietary data and domain expertise
Moat 3
Distribution
Reach and user relationships at scale
"Agents don't need UI."
ACT V
FROM THE INSIDE
What Failed
Top-Down Mandates
Forcing AI adoption through policy and directives. People resisted. Output quality was low.
What Worked
Bottom-Up Enablement
Unlimited compute + incentive alignment = 20x productivity. Let builders build.
100
Product Managers
600
Engineers
50
Designers
Weekly AI upskilling sessions. Tribal knowledge shared across teams. Everyone experiments, everyone learns.
"The most valuable managers will be the ones who figure out how to diffuse AI throughout the company."
ACT VI
THE NEW PLAYBOOK
The Principle
Design Workflows for Agents Like You Design Them for Humans
A PM workflow today: gather signal, synthesize context, draft spec, align stakeholders, ship decision. Each step can be an agent. The manager becomes the orchestrator.
Before
Manager assigns tasks to 8 humans. Coordinates via meetings. Reviews output manually.
After
Manager orchestrates 8 agents. Coordinates via prompts. Reviews output with evals.
01 PROMPTING Instruct precisely 02 EVALUATIONS Measure AI quality
The Shift
"Everyone is becoming a generalist."
When AI handles the depth, humans provide the breadth. The one-person billion-dollar startup is no longer a meme. It is the trajectory.
Action 1
AI Budget
Allocate a monthly budget to experiment with AI tools. Treat it as an investment.
Action 2
Build AI Fluency
Use AI tools daily. Not just ChatGPT. Learn prompting, agents, evaluations.
Action 3
Think in Workflows
Break your work into steps. Identify which steps agents can own entirely.
Action 4
Become a Generalist
Let AI handle the depth. You provide the breadth, the judgment, the taste.
So let me ask again...
01 / 34
← → or click to navigate