How AI Will Reshape the Firm and the Org Chart

Learning Outcomes

By the end of this study material, you should be able to:

1. Identify the three ways AI enters organizational structures
2. Explain how each scenario affects workers at different levels
3. Analyze the implications for red tape, worker agency, firm size, and pay
4. Evaluate the trade-offs and tensions within each scenario

---

Step 1: Foundation — Why This Matters

Before examining the scenarios, understand the core premise:

The question is not whether AI will enter firms, but where it will go in the hierarchy

Key Starting Assumption

• We are in a transition period, not a sudden total automation
• Firms will be hybrid — part human, part AI — for the foreseeable future
• How AI enters determines who gains power, who loses it, and how work changes

---

Step 2: The Three Scenarios

Think of these as three different answers to: "Where does AI sit on the org chart?"

---

Scenario 1: Beneath Us — "The Promoting"

[You]
  ↓
[AI agents] ← NEW

Core idea: Everyone keeps their job. AI fills a new layer below existing workers.

How it works:

Key insight:

• You do not lose your job
• You automate your past self — the old version of your tasks
• Your role shifts upward toward management, creativity, and judgment

Example:

Jace the accountant no longer clicks through spreadsheets. His AI agents do that. He now focuses on communication, coordination, and strategy — work his manager used to do.

Critical point about entry-level workers:

• The lowest rung disappears for humans
• Junior workers skip grunt work and immediately exercise judgment and creativity
• Loss: Traditional apprenticeship learning
• Gain: Bigger remit from day one

Why adoption starts at the bottom:

• Ground-level workers know the work best
• They can direct AI most effectively
• Managers unfamiliar with the details struggle to leverage AI tools

---

Scenario 2: Beside Us — "The Interleaving"

[Manager]
  ↓        ↓
[Human]  [AI team] ← NEW

Core idea: AI joins the org chart as a colleague, not just a subordinate tool.

How it works:

• Existing managers delegate entire functions to AI teams
• AI handles end-to-end tasks without constant human supervision
• Humans and AI teams report to the same manager

Example:

Sheila, a VP, gives an AI marketing team a budget and a product. The AI team interviews salespeople, runs surveys, prototypes visuals, and hands her a finished campaign on Monday. She approves it.

The "Escalator of Scope" concept:

Task → Job → Team → Firm
(narrow)              (broad)

• Harder to build AI that handles broader scope
• But easier for firms to adopt broader-scope AI
• A narrow tool requires retraining thousands of workers
• A broad AI team can be adopted by one executive with a single decision

The key asymmetry:

Important caution:

• Piecemeal AI beside every worker creates thousands of seams and handoffs
• This may actually slow adoption
• More capable, broader AI may ultimately be faster to adopt despite taking longer to build

---

Scenario 3: Between Us — "The Layering"

[Executive]
     ↓
   [AI] ← NEW
     ↓
[Workers]

Core idea: AI sits between executives and workers — coordinating, monitoring, advising.

How it works:

• AI acts as an always-on management consultant
• Scans communications, documents, and workflows
• Proactively surfaces problems and opportunities to executives
• Recognizes worker contributions that would otherwise go unnoticed

Example:

Asha's AI consultancy detects that sales promised a feature engineering isn't building — by reading Slack threads. It also notices Sarah quietly built useful infrastructure on her own initiative. The AI flags both to Asha, who can act on both immediately.

Why this is surprising:

• We assume automation climbs up from the bottom
• But management work lives mostly on computers and in the mind
• Some managerial functions may actually automate before the work they oversee

The "AI Consultancy" model:

• Traditional consulting: expensive, temporary, external
• AI consulting: cheap, permanent, internal
• No technical skill required from the executive — just conversation

Structural effect: Centralization

• Executives gain visibility into every corner of the organization
• The distance between C-suite and front-line shrinks dramatically
• Power concentrates at the top

The monitoring dimension:

• AI can understand real work, not just mouse clicks and green status dots
• Creates potential for genuine trust between employers and employees
• Could enable more remote work, flexible arrangements, and recognition of unconventional contributors

---

Step 3: The Four Big Questions

Now apply the scenarios to broader organizational consequences.

---

Question 1: More or Less Red Tape?

Current reality: Building is expensive → firms invest heavily in planning → lots of approval processes

AI changes this: Building becomes cheap → prototyping becomes easy → firms can "fire more, aim less"

Likely outcome:

• Fewer planning committees and sign-off layers
• More rapid iteration and experimentation
• Shift from memos to working prototypes as communication
• Individuals gain more power to act without permission

Tension:

The same AI that empowers individuals to prototype freely can also enforce more red tape from the top through the AI consultancy

---

Question 2: More or Less Worker Agency?

Two forces pulling in opposite directions:

Authors' view: On net, workers gain agency

• AI flags exceptions while leaving routine decisions with workers
• Workers become aware of business constraints and can be creative within them
• Bigger outputs mean more opportunities to demonstrate judgment

---

Question 3: Will Firms Shrink or Widen?

The Coase question: Why do firms exist at all? To reduce transaction costs, align incentives, and coordinate large projects.

AI could make firms larger:

• Executive can know a 10,000-person firm the way a founder knows a 20-person startup
• Better coordination across more people
• Vision can spread more broadly

AI could make firms smaller:

• Frictionless communication reduces need for vertical integration
• Better contract monitoring means firms can trust outside suppliers
• Less need to bring everything in-house

Likely outcome: Both

• Many smaller, nimble firms
• Some sprawling AI "holding companies" built around a common asset (factory, robot fleet, compute)
• The conglomerate model revived where a common mission justifies scale

---

Question 4: Will Pay Disparity Rise or Fall?

Two competing forces:

AI gives everyone the same great advice
        ↓
Less differentiation between workers
        ↓
Pay scales COMPRESS

versus

Each person manages a larger AI operation
        ↓
Small differences in judgment have bigger consequences
        ↓
Pay scales WIDEN

The key question:

Are workers mostly providing judgment, or following AI judgment?

---

Step 4: Synthesis — How the Scenarios Connect

These three scenarios are not mutually exclusive. They can coexist:

[Executive] ←— AI Consultancy (Layering)
     ↓
[VP/Manager] ←— AI team colleagues (Interleaving)
     ↓
[Individual Contributor] ←— AI agent workforce (Promoting)

A single firm could simultaneously have:

• AI beneath every worker
• AI beside certain teams
• AI advising executives from above

---

Step 5: Key Concepts to Remember

---

Quick Self-Test

1. Why might AI adoption start at the bottom of the org chart rather than the top?
2. What is the central asymmetry between building broader AI and adopting it?
3. How does the "AI consultancy" both increase and decrease worker agency?
4. Under what conditions would pay disparity increase with AI? When would it decrease?
5. Why might firms become smaller even as AI improves coordination?

---

One-Sentence Summary

AI will enter firms from multiple directions simultaneously — beneath workers as a labor force, beside them as colleagues, and between them as a management layer — reshaping hierarchy, agency, firm size, and pay in ways that depend critically on how capable and broadly adoptable the AI becomes.