New Laws of Robotics

Defending Human Expertise in the Age of AI

By Frank Pasquale • Cornell Law • 2020 • PROSE Awards Finalist

TL;DR:

Pasquale's four laws for AI governance: complementarity (augment, don't replace), authenticity (don't counterfeit humanity), cooperation (avoid zero-sum arms races), and attribution (traceability for accountability)—preserve human expertise and meaningful work.

Guide byAlbert Schaper

The Four Laws of Robotics

Unlike Asimov's laws (protecting humans from robots), Pasquale's laws govern how humans should develop and deploy AI.

1

Complementarity

Complementarity

AI should augment professionals, not replace them

Example: AI helps radiologists diagnose better, doesn't replace them

2

Authenticity

Authenticity

AI should not counterfeit humanity

Example: Chatbots must identify as AI, not pretend to be human

3

Cooperation

Cooperation

AI should not intensify zero-sum arms races

Example: International agreements on autonomous weapons

4

Attribution

Attribution

AI must be traceable to creators/controllers

Example: Clear responsibility chains for AI decisions

Intelligence Augmentation vs. Replacement

Same technology, radically different deployment choices.

Intelligence Augmentation (IA)

Goal: Make humans better at their work
Result: Human expertise becomes more valuable

Example:

AI diagnostic system → doctors make better diagnoses

Artificial Intelligence (AI)

Goal: Replace human performance
Result: Human expertise becomes obsolete

Example:

Fully automated diagnostic → radiologists unnecessary

High-Stakes vs. Low-Stakes Systems

Not all decisions should be algorithmic. Stakes determine whether human judgment is required.

Low-Stakes Systems

Algorithmic decisions acceptable

  • Entertainment recommendations
  • Ad targeting
  • Search results
  • Social media feeds
  • Shopping suggestions

High-Stakes Systems

Human judgment required

  • Criminal sentencing
  • Employment decisions
  • Healthcare diagnosis
  • Loan lending
  • University admissions

Professional Expertise Framework

Professional expertise includes more than technical knowledge:

What AI Can Help With:

  • Technical knowledge
  • Pattern recognition
  • Data analysis
  • Routine tasks

What Requires Humans:

  • Contextual judgment
  • Ethical reasoning
  • Relationships & trust
  • Wisdom & values

Frequently Asked Questions

What are Pasquale's Four Laws?▾
1) Complementarity: AI should augment, not replace. 2) Authenticity: AI should not counterfeit humanity. 3) Cooperation: AI should not intensify arms races. 4) Attribution: AI must be traceable to responsible parties. These laws govern humans developing AI, not robots themselves.
What's the difference between IA and AI?▾
Intelligence Augmentation (IA) makes humans better at their work—expertise becomes more valuable. Artificial Intelligence (AI) as replacement makes humans obsolete—expertise is eliminated. Same technology, radically different deployment choices.
Which decisions should remain human?▾
High-stakes decisions affecting human lives: criminal sentencing, employment, healthcare, lending, admissions. These require context, values, mercy, and accountability that algorithms cannot provide. Low-stakes decisions (recommendations, ads) can be algorithmic.
Why does expertise matter?▾
Professional expertise includes: technical knowledge (AI can help), contextual judgment (AI struggles), ethical reasoning (AI cannot do), relationships (AI lacks), and wisdom (AI doesn't have). Eliminating expertise eliminates what makes professions valuable.
How do we ensure AI augments rather than replaces?▾
Policy intervention: regulations incentivizing IA over AI, labor protections for professionals, democratic input into automation decisions, breaking corporate monopoly on technological choices. Automation isn't inevitable—it's a political choice.

Evaluate Tools with Pasquale's Framework

Apply the Four Laws

ChatGPT

Does it complement or replace expertise?

GitHub Copilot

Augments developers or replaces them?

Grammarly

Enhances writing or eliminates writers?

Explore All AI Tools Tools

Defend Human Expertise

AI Critical Thinking
How to Think About AI
Explore AI Tools

Key Insights: What You've Learned

1

Pasquale's four laws for AI governance provide essential principles: complementarity (augment human expertise, don't replace it), authenticity (don't counterfeit humanity), cooperation (avoid zero-sum arms races), and attribution (ensure traceability and accountability)—these laws preserve human dignity and meaningful work.

2

Apply these laws by distinguishing high-stakes from low-stakes AI systems, preserving human expertise in critical domains, avoiding deceptive AI that mimics humans, and ensuring transparency and accountability—effective AI governance requires balancing innovation with human values.

3

Build AI systems that enhance rather than diminish human agency: use AI for augmentation in high-stakes contexts, maintain human oversight, preserve meaningful work, and design systems that respect human expertise—the goal is intelligence augmentation, not replacement.