The Overlooked Opportunity

The problem of deferred work isn't unique to technology—it's a universal feature of every complex system operating under resource constraints. From Canada to Australia, from Britain's NHS to Germany's crumbling bridges, healthcare, legal systems, and public infrastructure all accumulate vast backlogs of maintenance work that never gets prioritized.

Healthcare's Administrative Burden

In healthcare, the crisis is global. Canadian physicians spend 18.5 million hours annually on unnecessary administrative tasks—equivalent to 9,250 full-time positions that could provide patient care instead ¹. Britain's NHS waiting lists reached 6 million people, up from 4.4 million pre-pandemic, with elective care standards unmet since 2016 ². The administrative documentation burden drives physician burnout while creating "note bloat" that obscures rather than clarifies patient information. Yet healthcare systems lack capacity to rationalize these processes while managing immediate patient demand.

Legal System Backlogs

Legal systems face parallel pressures. The Court of Justice of the European Union struggles with cases averaging over 17 months to process, with 1,206 cases pending as of December 2024 ³. The Council of Europe created a comprehensive "Backlog Reduction Tool" in 2023 specifically because judicial systems across member states cannot keep pace with incoming cases ⁴. As of January 2025, 650 European Court of Human Rights judgments remain unimplemented across EU member states—each representing unresolved human rights issues—with 45.7% of the past decade's judgments still pending ⁵. The maintenance work of justice—case review, precedent research, implementation of reforms—accumulates while courts struggle with immediate caseloads.

Infrastructure Investment Gaps

Public infrastructure reveals the global scale. Germany faces an infrastructure investment gap estimated at €400-500 billion, with Deutsche Bahn alone requiring €165 billion by 2034 for railway network renewal ⁶. Australia's Infrastructure Australia reports a mounting maintenance backlog driven by "historical underspending on preventative maintenance, short budgetary cycles, and inadequate reporting," warning that deferred maintenance costs four times as much to address later ⁷. Europe faces a $2 trillion infrastructure gap by 2040, even as the EU's Connecting Europe Facility received €9.5 billion in project requests against only €2.5 billion available ⁸.

The pattern is identical across sectors: critical maintenance work that everyone acknowledges needs doing, that will cost dramatically more if deferred further, but that never gets prioritized because immediate demands consume all available resources.

The AI Opportunity

This is where AI could deliver genuine societal value—not by replacing human judgment on critical decisions, but by addressing the maintenance backlogs that humans don't have capacity for. And crucially, managing these AI-driven maintenance projects could provide exactly the kind of entry-level work we need to preserve: real responsibilities with real impact, building deep understanding of domain knowledge, all while producing valuable output under appropriate supervision.

Imagine entry-level legal staff managing AI systems that help process court documentation backlogs, learning case law and procedural knowledge while keeping human judgment on substantive legal questions. As they review AI-generated case summaries and identify patterns, they build the deep domain expertise that prepares them for substantive legal work—learning not by replacing judgment, but by exercising it at scale under supervision.

Entry-level healthcare administrators could use AI to rationalize medical record systems, gaining understanding of clinical workflows while freeing physicians to focus on patients. Junior civil engineers might direct AI analysis of infrastructure condition assessments, building expertise in municipal systems while helping clear maintenance backlogs. In each case, the work is real, the impact is measurable, and the learning is foundational—exactly what entry-level positions should provide.

True Productivity vs. Wealth Extraction

This is how true productivity works: maintaining employment while increasing value creation. The alternative—using AI to eliminate workers and transfer their productivity gains to capital owners—isn't productivity. It's wealth extraction. When AI replaces a worker, the value that worker created doesn't disappear—it flows to whoever owns the AI system. If that worker would have spent their wages in the economy while the AI owner pockets the gains or reinvests in eliminating more workers, we've reduced economic velocity while concentrating wealth. That's the opposite of productive.

The productivity gains from parallelization are real. Where tasks are repetitive with small variations, AI agents can handle them simultaneously rather than sequentially, provided each has necessary resources. This isn't about eliminating humans; it's about freeing human capacity for analysis, creativity, and strategic thinking—the work that drives innovation—while tackling the maintenance backlogs that accumulate in every complex system.

The Diversity Imperative

Research shows that diversity improves innovation and productivity ⁹. Cultural diversity specifically enhances corporate AI adoption by strengthening innovation capabilities ¹⁰. Organizations in the top quartile for gender diversity on executive teams are 25% more likely to achieve above-average profitability ¹¹.

But here's a critical caveat: recent analysis of 40+ million academic papers reveals that while AI users publish three times as many papers and receive nearly five times as many citations, "AI-heavy research occupies a smaller intellectual footprint, clusters more tightly around popular, data-rich problems, and generates weaker networks of follow-on engagement between studies" ¹². Individual productivity increases, but intellectual diversity decreases.

This suggests that simply deploying AI without deliberate attention to diversity—of ideas, approaches, and perspectives—risks creating a paradox: more output, narrower thinking. The solution isn't to avoid AI, but to structure its deployment to enhance rather than constrain intellectual exploration, to address genuine backlogs while maintaining the human pathways that build expertise and judgment.

References

This article draws on research current as of January 2026.