Why Building AI Era Talent Pipelines Requires More Than Corporate Training Programs

The Talent Challenge Facing Business Leaders

The talent challenge facing business leaders has reached an inflection point. Artificial intelligence is not simply automating routine tasks; it is fundamentally restructuring how work gets done, what skills matter, and how organizations must think about developing human capital. Beth Cobert, president of affiliates and strategic partnerships at Strada Education Foundation and former senior partner at McKinsey, argues that the traditional playbook for workforce development no longer suffices. Organizations cannot address this challenge in isolation.

The Core Thesis: Partnerships as Survival Strategy

Cobert's central argument rests on a simple premise: AI's workforce impact is too broad and too rapid for any single organization to address alone. She advocates for "systems change" rather than programmatic interventions—building infrastructure that enables multiple stakeholders to access reliable information and create sustainable pathways from education to employment.

The logic is compelling. When Cobert led programs at the Markle Foundation, her team spent weeks manually researching which training programs delivered good outcomes for working adults. This approach could help 20 people, not 20,000. The solution? Strada's affiliate CredLens is creating a national data trust with verifiable credential outcomes, allowing career coaches and employers to access this information instantly.

This represents a fundamentally different approach to workforce development—infrastructure over intervention, ecosystems over programs. The question is whether this model can scale quickly enough to meet the moment.

The Entry-Level Jobs Dilemma

One of the most pressing issues Cobert addresses is the disappearing entry-level role. As AI handles basic tasks, traditional career ladders are losing their bottom rungs. A 2025 New York Times article reported computer science graduates from prestigious schools seeking work at retail chains, unable to find the entry-level positions that existed just years earlier.

Cobert acknowledges this trend but suggests we haven't fully grasped its implications. If entry-level positions vanish, how do workers build the judgment and experience needed for decision-making roles? Her answer involves integrating learning into work more seamlessly and creating faster communication loops between employers and educators.

Yet this response may underestimate the severity of the problem. Brookings Institution research shows entry-level hiring in tech sectors declined 23% between 2022 and 2024. The "missing rung" phenomenon creates a cascade effect: without entry experience, mid-career advancement becomes harder, and organizations lose their traditional talent development pipeline.

The partnership model Cobert advocates could address this, but only if employers commit to creating new entry points. The Denver advanced manufacturing institute she cites—a collaboration between Community College of Denver, Metro State University, University of Colorado Denver, and employers like Lockheed Martin—shows this is possible. Students can transfer credits between institutions while gaining employer-integrated experience. But replicating such arrangements across sectors and geographies requires coordination that most regions lack.

The Skills-Based Hiring Promise and Reality

Cobert is a passionate advocate for skills-based hiring, sharing a story about a lens manufacturer that couldn't find qualified machine operators until the CEO hired a sushi chef. The company realized it needed attention to detail, manual dexterity, and patience—skills found in food preparation workers and manicurists, not just engineers.

This narrative powerfully illustrates how credential fixation limits talent pools. Research from Harvard Business School and Accenture identified 27 million "hidden workers" in the US—people screened out by automated systems focused on degrees and employment gaps rather than actual capabilities. Skills-based hiring could unlock this talent.

However, implementation proves far more difficult than advocacy. IBM's "New Collar" initiative eliminated degree requirements for technical roles, focusing on skills instead. While this expanded candidate pools, the company encountered significant challenges: managers defaulted to familiar credentials, skills assessment tools produced inconsistent results, and determining equivalent experience proved subjective.

The lens manufacturer story also glosses over a critical detail: the CEO personally discovered this talent through conversation. Most organizations hire at scale through processes designed for efficiency, not for uncovering hidden potential. Moving from anecdote to systematic practice requires assessment infrastructure most companies don't possess.

Cobert suggests AI might actually help here, enabling better skills matching. But this assumes AI tools can evaluate soft skills like "getting things done" that she identifies as crucial. Natural language processing struggles with these nuanced human capabilities, and over-reliance on AI screening might paradoxically exclude exactly the unconventional candidates skills-based hiring aims to include.

Investing in Incumbent Workforces

One of Cobert's strongest arguments addresses incumbent workers. As AI automates basic tasks, she argues, organizational value increasingly lies in judgment—understanding what works for your company and how to deliver value to customers. This knowledge resides in current employees' "heads and hearts." Without investment in upskilling them, organizations risk losing institutional knowledge in favor of generic, AI-driven answers.

This perspective challenges the assumption that AI makes human workers less valuable. Instead, it suggests AI raises the value of experienced workers who can guide AI tools with contextual understanding. A junior employee might use AI to draft a client proposal, but does the draft reflect the client relationship's nuances? Does it align with the company's strategic positioning? Experienced workers bring judgment that AI cannot replicate.

Research supports this view. MIT's Work of the Future studies found that AI's impact depends heavily on how organizations deploy it. Companies that use AI to augment worker capabilities see productivity gains without displacement. Those that use AI primarily for labor substitution often lose valuable tacit knowledge.

AT&T's Workforce 2020 initiative offers a relevant case study. Facing network technology transformation, the company invested $1 billion to retrain over 100,000 employees. The program succeeded in maintaining workforce capability through major technological change, but it required CEO commitment, significant resources, and years of sustained effort. Most organizations lack these preconditions.

Cobert expresses confidence that investing in employees pays off and contradicts the belief that trained workers leave. Her experience may differ from broader evidence. During economic uncertainty, training budgets often face cuts. The Conference Board reports that learning and development spending decreased 14% in 2023-2024 as companies prepared for recession. When businesses most need workforce adaptation, investment often declines.

The Systems Change Imperative

The most ambitious aspect of Cobert's argument involves "systems change"—building infrastructure that enables sustainable, scaled solutions rather than replicating programs one organization at a time.

She describes the education-to-employment landscape as "a set of individual organizations and institutions not acting in a coordinated way." True ecosystems require clear interaction patterns and connections, which most regions lack. The Denver advanced manufacturing example shows what's possible: institutions sharing facilities, credits transferring seamlessly, and employers integrated throughout.

But creating these ecosystems requires addressing coordination challenges that have stymied reformers for decades. Educational institutions operate on academic calendars with accreditation requirements that change slowly. Employers need workforce skills that shift rapidly. Governments provide funding but often through categorical programs that don't align. Students and workers navigate this complexity without reliable information about which pathways lead where.

CredLens, the data trust Cobert describes, attempts to solve one piece: providing verifiable credential outcomes. This matters enormously. Workers choosing training programs need to know which lead to good jobs. Employers need to understand what different credentials signify. Currently, this information is scattered, inconsistent, and often outdated.

Singapore's SkillsFuture program offers an international comparison. The national system connects training providers, credential recognition, and employer input, with government subsidy. It provides the infrastructure Cobert envisions, but it required centralized governance, substantial public investment, and cultural consensus about workforce development's importance. Replicating this in the US fragmented education and training landscape presents significant challenges.

The question becomes: Can bottom-up partnership approaches achieve systems change, or does this require top-down policy intervention? Cobert's examples suggest voluntary collaboration can work locally, but scaling nationally may require policy frameworks that create incentives for participation.

What Skills Will Matter?

Cobert takes a nuanced position on skills, arguing against purely technical training. She maintains that foundational capabilities—communication, relationship building, problem-solving, and the ability to "get things done"—remain valuable even as AI handles technical tasks.

This perspective aligns with research on AI's complementary relationship with human skills. MIT economist David Autor argues that AI strengthens demand for workers who can exercise judgment, creativity, and interpersonal skills. The challenge lies in credentialing and demonstrating these capabilities.

Cobert offers a powerful example: a single parent working two jobs while completing education demonstrates resourcefulness, time management, and determination. Yet traditional hiring processes may screen out this person due to employment gaps or lack of conventional credentials. She argues we need better ways to recognize evidence of these capabilities.

This is where skills-based hiring and systems change intersect. If CredLens can verify not just technical training outcomes but also soft skill development, it creates new possibilities for talent recognition. However, assessing capabilities like adaptability and collaborative problem-solving presents measurement challenges that technical skills don't.

The World Economic Forum's Future of Jobs Report identifies critical thinking, creativity, and emotional intelligence as top skills for 2025 and beyond. But how do educational institutions teach these? How do workers demonstrate them? How do employers assess them at scale? These questions lack clear answers.

Cobert suggests AI itself might help workers develop and demonstrate skills in new ways. Students are already asking how to use AI in daily tasks and what skills to develop. This creates opportunities for educational institutions to focus on AI fluency—not just using tools but understanding when AI's recommendations make sense and when human judgment should override them.

The Implementation Challenge

When asked how organizations should start transformation, Cobert offers pragmatic advice: find a real need with willing leadership, achieve early wins, and build a coalition that attracts others. She notes that this approach converts about three-quarters of an organization, with the remainder requiring direction once success is proven.

This change management wisdom is sound, but it may underestimate the barriers specific to workforce development partnerships. Unlike internal organizational change, education-employment ecosystems involve multiple independent entities with different incentives, timelines, and cultures.

Year Up, a nonprofit that bridges education and employment for young adults, demonstrates both the potential and limitations of the partnership approach. The organization achieves 85% job placement rates by combining technical training, professional skills development, and employer partnerships with six-month internships. It's a proven model, yet Year Up operates primarily in major metropolitan areas and has struggled to scale to mid-sized cities where coordination is harder and employer density lower.

The coordination challenge intensifies with AI's rapid pace. Cobert emphasizes that educational institutions must accelerate communication with employers about changing needs. But academic program development typically requires two to three years from conception to implementation—far too slow for AI-driven changes. Faster adaptation might require more modular credentials, work-integrated learning, and employer-sponsored training that educational institutions validate rather than develop from scratch.

The Equity Dimension

Cobert's focus on "those who have faced the biggest barriers" is central to Strada's mission. She argues that systems change must expand opportunity for all, not just those already connected to institutional networks. The single parent example illustrates how traditional pathways exclude capable people.

Yet partnership approaches risk reinforcing existing inequities. Organizations tend to partner with institutions and communities where they already have relationships. The Denver advanced manufacturing institute serves its region, but similar models haven't emerged in communities with weaker institutional capacity or employer density. Rural areas, smaller cities, and communities with fewer anchor institutions may lack the preconditions for partnership-based approaches.

This suggests that voluntary partnerships, while valuable, cannot substitute for policy interventions that ensure equitable access. Federal workforce development funding, state coordination of educational institutions, and employer incentives for training investment all matter for reaching populations that market-driven partnerships might miss.

Research on workforce development partnerships shows that successful collaborations tend to emerge in regions with pre-existing civic infrastructure, employer associations, and institutional trust. Building this capacity in communities that lack it requires sustained investment in organizational development, not just programmatic funding.

Beyond AI: Demographic and Economic Factors

Cobert wisely notes that AI is not the only labor market force demanding attention. Demographic changes matter enormously. The US working-age population is growing more slowly than in previous decades, with fertility rates declining and immigration reduced from historical levels. This creates workforce scarcity that makes talent development more critical.

The care economy presents particular challenges. Healthcare, childcare, and eldercare jobs require human skills and are growing rapidly, yet many pay poorly and lack clear advancement pathways. If AI eliminates entry-level roles in knowledge work while expanding demand in care work, we need strategies to make care careers economically viable and professionally rewarding.

Cobert also addresses the changing psychological contract between employers and workers. As employment has become more transactional, workers reasonably ask what employers will invest in them. Organizations that cannot answer this question compellingly will struggle to attract and retain talent, regardless of compensation.

This points to a broader transformation in how organizations think about human capital. The industrial model treated workers as costs to minimize. The knowledge economy model invested in high-skilled workers while treating others as replaceable. The AI era may require a more universal commitment to worker development—not from altruism but from recognition that organizational success depends on workers who can guide, interpret, and contextualize AI capabilities.

A Critical Assessment

Cobert's arguments for partnership-based, systems-level workforce development are compelling, particularly for regional ecosystems with strong institutional capacity. The examples she cites—Denver's manufacturing institute, MiraCosta College's biotech program—demonstrate that collaboration can create pathways that benefit students, workers, and employers.

However, several concerns warrant attention:

• First, the pace problem persists. Even with stronger partnerships, educational institutions change more slowly than AI capabilities evolve. This mismatch may require shifting more learning into the workplace itself, with educational institutions validating and credentialing rather than exclusively delivering training. Corporate investment in worker development becomes more critical, raising questions about employers' willingness to make these commitments.
• Second, the measurement challenge remains unresolved. Skills-based hiring requires reliable skills assessment. While CredLens can provide outcome data about credential programs, assessing individual capabilities—particularly soft skills—proves difficult at scale. Without better assessment infrastructure, skills-based hiring risks remaining more rhetoric than reality.
• Third, equity gaps may widen. Partnership approaches work best where institutional capacity and employer density already exist. Ensuring that rural communities, smaller cities, and marginalized populations benefit requires policy interventions beyond voluntary collaboration.
• Fourth, the economic uncertainty factor cannot be ignored. Cobert's optimism about employer investment in workers runs counter to cyclical patterns where training budgets face cuts during downturns. Making workforce development sustainable requires mechanisms that maintain investment through economic cycles.
• Finally, the governance question looms large. Who coordinates these partnerships? Who ensures that employer needs don't completely dictate educational priorities? How do we balance immediate skill needs with broader educational missions? These questions lack clear answers.

A Path Forward

Despite these concerns, Cobert's core insights remain valid. The AI-era talent challenge is too large and too rapid for organizations to address alone. Partnerships between employers, educational institutions, and community organizations offer the most promising path forward. Several principles should guide implementation:

• Start with regional ecosystems. Geographic proximity facilitates the relationship-building and coordination that partnerships require. Cities and metropolitan regions can convene stakeholders more effectively than national initiatives.
• Invest in infrastructure, not just programs. Data trusts like CredLens, shared facilities like Denver's manufacturing center, and credential translation services create lasting capability rather than one-time interventions.
• Focus on portability and transparency. Workers need credentials and skills documentation that transfer across employers and regions. Transparent outcome data helps workers make informed choices and creates accountability for training providers.
• Embed learning in work. Rather than separating education and employment, create pathways that integrate both—apprenticeships, internships, work-based learning, and employer-sponsored development that educational institutions validate.
• Maintain investment through cycles. Policy mechanisms, industry associations, and labor-management agreements can help sustain workforce development investment when economic pressures create short-term incentives to cut.
• Design for equity from the start. Partnerships should explicitly address how they will reach workers facing significant barriers, not assume that benefits will naturally spread.

Cobert concludes her interview with optimism rooted in specific examples—students at Pueblo Community College opening healthcare careers, diverse summer interns at Strada excelling in their roles. This grounded optimism recognizes real challenges while seeing demonstrated solutions.

Business leaders face a choice: retreat to narrow, short-term workforce strategies focused on their immediate needs, or engage in the harder work of building regional talent ecosystems that serve broader purposes while meeting organizational requirements. The former might seem easier but risks leaving organizations without the talent they need. The latter requires patience, coordination, and investment but creates sustainable capability.

The AI era does not permit standing still. Technologies evolve too rapidly, skills shift too quickly, and competitive advantage increasingly depends on organizational learning capacity. Whether business leaders embrace partnership approaches or not, the workforce development system will transform. The question is whether that transformation will be chaotic and inequitable or intentional and inclusive.

Cobert's vision of connected ecosystems replacing fragmented institutions offers a compelling alternative to both complacent incrementalism and dystopian displacement scenarios. Realizing this vision requires business leaders to recognize that talent development is not someone else's problem to solve. It requires sustained commitment, even when quarterly pressures argue otherwise. And it requires humility about the limits of what any organization can accomplish alone.

The organizations that thrive in the AI era will be those that invest in their people, partner with their communities, and recognize that workforce development is not a cost to minimize but a capability to build. This is not a comfortable message for leaders accustomed to controlling their destiny, but it may be the most important lesson the AI transformation has to teach.