Left solely to a shareholder-driven marketplace, AI has too often become a tool for mass layoffs rather than shared progress. Thousands of engineers and other skilled employees have already been dismissed by the Tech Giants and Hyperscalars as they pursue efficiency gains measured as higher quarterly returns. Yet another path is possible. By broadening the definition of value beyond shareholders alone to include employees, citizens, and the long-term health of society, AI can be developed to augment human capability instead of replacing it. Building that future will require innovation not only in technology, but also in economic thinking, corporate governance, and public policy.

Recursive job displacement refers to a self‑reinforcing sequence of task automation:

1. AI automates a set of routine or structured tasks.
2. Workers shift into remaining tasks that are less protected or lower paid.
3. As AI improves, those remaining tasks become the next targets for automation.
4. The cycle repeats, reducing the overall “task space” available to human labor.

This differs from traditional automation because the cycle is continuous, faster, and increasingly affects cognitive tasks.

Potential Benefits

• Higher productivity: AI can increase output and efficiency across sectors.
• Cost savings: Firms can streamline operations and reduce administrative burdens.
• New task creation: Oversight, design, ethics, and human‑machine coordination roles can emerge.

Key Risks

• Task hollowing: Middle‑skill roles shrink, pushing workers downward.
• Accelerated displacement: Generative AI targets analytical and decision tasks.
• Inequality pressures: Gains may concentrate among firms and capital owners.
• Weaker bargaining power: Workers lose leverage as task content erodes.

Some Policy Approaches Under Consideration

1. China’s Emerging Legal Approach: Protect Displaced Workers

Recent legal decisions and draft regulations in China emphasize that employers cannot rely on automation as a blanket justification for layoffs. Courts have required:

• Demonstrated necessity for automation‑related terminations
• Evidence of attempted retraining or reassignment
• Enhanced compensation when displacement is unavoidable

This model treats technological displacement as a regulated employment event, with obligations placed on firms to mitigate harm.

2. Traditional Industrial Approach: Market Absorption

Historically common in many advanced economies:

• Firms automate and reduce headcount.
• Workers are expected to seek new employment independently.
• Retraining is voluntary and often fragmented.

This approach maximizes managerial flexibility but intensifies recursive displacement, as workers repeatedly re‑enter the labour market with fewer transferable tasks.

3. Internal Downgrading: Retain Workers, Reduce Job Quality

Some organizations avoid layoffs by:

• Reassigning displaced workers to lower‑skill, lower‑wage roles
• Maintaining employment while reducing autonomy, pay, or career progression

This slows unemployment growth but compresses wages and skill levels, reinforcing downward mobility within firms.

4. Augmenting Labour: Expanding Human Capability Rather Than Replacing It

A growing policy priority is ensuring that AI augments workers rather than substitutes for them. This requires structured training, redesigned workflows, and incentives for firms to adopt complementary rather than replacement AI systems.

AI can enhance human performance by:

• Automating low‑value tasks, freeing workers to focus on judgment, relationship‑building, creativity, and problem‑solving.
• Acting as a decision-support tool, providing analysis, summaries, and scenario modelling that improve accuracy and speed.
• Expanding individual capacity, enabling one worker to handle more complex portfolios (e.g., caseworkers, analysts, technicians).
• Improving quality and consistency, especially in documentation, compliance, and customer service.
• Supporting continuous learning, with AI copilots that coach workers in real time.

In this model, AI becomes a force multiplier, not a replacement.

Policy Levers to Support Complementary Use

• National AI‑literacy and task‑specific training programs tied to real occupations.
• Subsidies or tax credits for firms that adopt AI systems demonstrably designed to augment workers.
• Certification frameworks for “human‑in‑the‑loop” AI tools that preserve worker autonomy.
• Work redesign grants that help organizations restructure roles around human‑AI collaboration.
• Public sector pilots that model best practices in augmentation rather than substitution.

This approach aims to expand the human task frontier, counteracting the recursive displacement cycle.

Implications for Policymakers

The four approaches reflect different philosophies of how societies should manage technological change. Key considerations include:

• Economic stability: Whether displacement is absorbed by firms, markets, or the state.
• Worker security: Whether protections are proactive (China model) or reactive (traditional model).
• Task design: Whether organizations redesign work to preserve meaningful human roles.
• Skill development: Whether workers are equipped to use AI as a complement rather than a competitor.

Governments evaluating policy options may need to balance innovation incentives with safeguards that prevent long‑term erosion of human task space.