How Gen AI Could Change the Value of Expertise | Harvard Business Review (2025)

EMPLOYABILITY & LEARNING CULTURE

How Gen AI Could Change the Value of Expertise | Harvard Business Review (2025) | The AI revolution isn't just changing job requirements—it's fundamentally reshaping learning curves. Discover how generative AI simultaneously creates barriers and bridges to expertise in ways that will transform recruitment, promotion paths, and entire organisational structures.

📊 DID YOU KNOW?

👀 DID YOU SEE?

The article's striking "Earning Curves by Occupation" visualisation reveals a stark divide in how expertise is valued across professions. Software developers, actuaries, and credit analysts show steep upward curves where top performers earn over $150,000 —more than double their entry-level counterparts. Meanwhile, cashiers and similar roles display nearly flat lines with minimal wage growth regardless of experience. This visual perfectly illustrates which careers face AI disruption at the entry level (the steep curves) versus those where AI might democratise access (the flat curves), allowing readers to instantly grasp where their profession might fall in this new paradigm.

✨ OVERVIEW

The Burning Glass Institute and Harvard Business School Project on Managing the Future of Work have uncovered a surprising bifurcation in how generative AI is reshaping expertise acquisition. Their research reveals that about 12% of U.S. workers are in roles where AI will likely automate away entry-level positions, whilst 19% are in fields where AI will make technical skills more accessible to newcomers. This dual impact will affect roughly 50 million jobs in the coming years, forcing a complete rethink of organisational structures and talent pipelines. The traditional pyramid-shaped organisation with many junior staff supporting fewer seniors is transforming into a diamond shape, fundamentally altering how expertise is developed, valued, and leveraged—creating crisis and opportunity for organisations that can adapt quickly.

🧩 CONTEXT

Professional advancement has followed predictable learning curves for decades, from novice to expert shaping everything from salary structures to organisational hierarchies. In high-value fields like finance, consulting, and technology, significant expertise gaps between juniors and seniors created natural career ladders and wage premiums for experience. These time-worn progressions established the fundamental blueprint for how companies hire, train, and promote talent.

Now, generative AI is disrupting this model in both directions. In some fields, it's automating the foundational tasks that historically provided newcomers with on-the-job learning, creating a "missing rung" on career ladders. Simultaneously, in other fields, AI democratises access to technical expertise that previously required years of specialised training. This seismic shift is happening just as employers face talent shortages and demographic challenges, creating a perfect storm of workforce transformation.

🔍 WHY IT MATTERS

↳ The expertise divide is widening—Research identifies 100 occupations where gen AI will likely raise entry barriers, affecting 17.8 million workers (about 12% of the U.S. workforce). These roles—including project management specialists, training managers, and financial risk specialists—show steep learning curves where experienced workers command significantly higher salaries. As AI automates entry-level tasks in these fields, the traditional pathways for developing future talent are being dismantled before our eyes.

↳ Traditional organisational structures are becoming obsolete—Current pyramid models with 5:1 ratios of juniors to seniors will rapidly shift toward 2:1 or even flatter structures. This will not just change headcounts but fundamentally transform how work happens. With fewer management layers, decision-making will accelerate, but companies risk creating critical talent bottlenecks without the traditional pipeline of trained juniors advancing to senior roles.

↳ New doors are opening for previously excluded talent—For 28.6 million workers (nearly 20% of the workforce) in fields like data warehousing, construction management, and network administration, AI is lowering barriers by handling technical aspects that once required specialised training. This represents a historic opportunity to diversify talent pools and address chronic skills shortages by bringing in previously excluded workers despite having relevant capabilities in other areas.

💡 KEY INSIGHTS

↳ Learning curve steepness predicts AI's barrier effect—The researchers discovered that occupations where experienced workers out-earn juniors by the most significant margins (steep learning/earning curves) are precisely where AI will most likely automate entry-level positions. In these fields, such as software development, where top performers earn $150,000+ compared to entry-level salaries of $55,000, AI tools like Copilot are already eliminating traditional junior roles. This pattern appears consistently across industries from finance to creative fields.

↳ Explicit knowledge barriers fall whilst implicit knowledge becomes more valuable—Gen AI excels at making codified, explicit knowledge (learned from textbooks) more accessible. In contrast, tacit, implicit knowledge gained through experience remains AI-resistant. This shifts the expertise premium toward company-specific knowledge and less codifiable skills. For example, a construction manager no longer needs to memorise building codes (AI handles that) but still needs on-site experience managing complex projects and stakeholder relationships.

↳ Diamond-shaped organisations require completely new talent models—With fewer entry-level positions and a continued need for senior expertise; organisations must develop entirely new approaches to talent. The research shows companies will likely need to recruit laterally from adjacent fields, create accelerated learning programmes to rapidly develop expertise, and invest heavily in retaining experienced staff. This represents a shift from the traditional "churn-and-burn" talent management model toward long-term retention and continuous skill development.

🚀 ACTIONS FOR LEADERS

↳ Redesign your organisational structure for the AI era—Audit your current role ratios between junior and senior positions, especially in fields with steep learning curves. Create transition plans to shift from pyramid to diamond structures by identifying which entry-level tasks can be automated and how to preserve necessary learning experiences. Set clear 3-year targets for new organisational shapes with metrics to track progress.

↳ Build rapid expertise development programmes—Traditional on-the-job learning will no longer suffice. Develop simulation-based training that compresses years of experience into months, focusing on company-specific knowledge that AI can't easily replicate. Follow the example of firms creating "digital twins" of complex systems where newcomers can gain virtual experience without risk. Allocate 15-20% of your L&D budget to accelerated expertise programmes.

↳ Create new cross-functional career pathways—With fewer traditional promotion opportunities, design lateral career movements that build versatile expertise. Map skills that transfer between functions and create formal rotation programmes. PwC and leading firms require at least two to three cross-functional experiences before considering senior leadership. Document your new career paths explicitly and train managers to guide staff through non-traditional advancement opportunities.

🔗 CONCLUSION

The gen AI revolution isn't just changing job requirements—it's fundamentally reshaping how expertise is developed, valued, and deployed. Organisations face a profound restructuring as traditional learning curves are redrawn and pyramid structures flatten into diamonds. This transformation creates threats and opportunities: whilst established talent pipelines are disrupted, new doors open for previously excluded workers in fields where AI can augment their capabilities.

Forward-thinking leaders will view learning curves as not fixed realities but strategic variables they can influence. By creating new expertise development models, redesigning organisational structures, and embracing AI-augmented learning, companies can build more flexible, diverse workforces capable of thriving amid constant technological change. Those who simply react to these shifts rather than proactively reshaping their talent approaches risk losing the expertise needed to remain competitive in an increasingly AI-powered world.

🎯 KEY TAKEAWAY

The true competitive advantage in the AI era isn't the technology but how organisations redesign their expertise development models. Those who master accelerated learning curves and diamond-shaped talent structures will outperform those who cling to traditional hierarchies and career paths.

Explore Our Themed Collections: Uncover a curated selection of current and past articles on our website.

Join 10,000+ Professionals: Follow me on LinkedIn for daily insights.