AI tools capture frontline tacit knowledge, closing Polanyi’s paradox for businesses

Polanyi’s paradox meets AI on the frontline

TechRadar reports that businesses are confronting Polanyi’s paradox—the idea that "we can know more than we can tell"—by deploying AI tools to capture and structure tacit knowledge held by frontline workers in warehouses, retail, and logistics. This paradox, first articulated by polymath Michael Polanyi, highlights how critical but undocumented expertise often resides with employees who perform hands-on tasks.

Up to 85% of a company’s knowledge is tacit, existing in the form of informal insights, shortcuts, and problem-solving methods rather than official manuals or standard operating procedures (SOPs). This knowledge is unevenly distributed, and its loss—when employees leave or retire—can disrupt operations and slow down onboarding and training.

AI knowledge agents as a solution

AI tools, particularly knowledge agents, are being used to crowdsource and structure this undocumented expertise. According to TechRadar, these agents are trained on existing SOPs and onboarding materials but also learn from interactions with frontline workers. By surfacing gaps in guidelines or training programs, they create a feedback loop that dynamically updates institutional knowledge.

Multimodal AI agents further enhance this process by processing audio, text, images, and video. This allows workers to contribute and access knowledge in formats that suit their roles—whether through voice queries, annotated images, or video demonstrations. The result is a more connected and adaptive frontline workforce, where knowledge flows in both directions: from workers to systems and back.

Why AI implementation succeeds—or fails

The success of AI adoption in capturing tacit knowledge hinges on more than just technology. TechRadar cites data showing that businesses partnering with AI providers achieve a 66% success rate in implementation, compared to 33% for those relying solely on internal development. The gap underscores the complexity of integrating AI into existing workflows and the value of external expertise.

A culture of experimentation and executive sponsorship is equally critical. MIT economist David Autor, quoted in the report, notes that tasks requiring flexibility, judgment, or common sense—those least likely to be codified—are the most challenging to automate. AI’s value, therefore, lies not just in algorithms but in rethinking workplace culture to encourage adoption and continuous learning. Stanford’s Digital Lab emphasizes that executive leadership and a willingness to embrace failure are key drivers of AI success.

The corporate knowledge economy

The shift toward AI-powered knowledge management is creating what TechRadar describes as a "connected frontline knowledge economy." In this model, SOPs act as the nervous system, while AI tools serve as the brain—crowdsourcing insights from workers and routing them back into training programs and operational guidelines. This two-way exchange helps businesses close knowledge gaps, improve job satisfaction, and reduce reliance on informal gatekeeping.

For industries like retail, warehousing, and logistics, where employee turnover is high and training demands are constant, this approach could improve efficiency and retention. However, the report warns that without the right cultural and leadership support, even the most advanced AI tools risk underdelivering.

Link to AI governance challenges

This story aligns with earlier Lazyfounder reporting on the broader challenges of AI implementation, particularly the governance gaps identified in UK businesses. While UK firms are investing heavily in AI, many struggle to measure ROI or achieve meaningful outcomes due to fragmented responsibility and lack of cost visibility. The emphasis here on executive sponsorship and cultural change mirrors findings that AI success depends as much on organizational readiness as on technology.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

For founders and operators, this story highlights a practical application of AI that goes beyond automation: capturing and leveraging tacit knowledge to strengthen workplace culture, training, and operational resilience. The 66% success rate for businesses partnering with AI providers suggests that external expertise may be a faster path to value than building solutions in-house—especially for companies with high turnover or complex frontline operations.

The focus on multimodal AI also signals an opportunity for startups developing tools that integrate voice, video, or image-based interactions. If frontline workers can contribute knowledge in ways that feel natural to them, adoption is likely to be higher, and the quality of data captured may improve.

However, the story underscores that AI is not a plug-and-play solution. The role of executive sponsorship and a culture of experimentation echoes broader research on AI governance, reinforcing that leadership buy-in and a willingness to iterate are non-negotiable for success. For founders, this means pitching AI not just as a cost-saving tool but as a catalyst for organizational change—one that requires alignment across teams, from HR to operations.

Related on Lazyfounder

• UK AI Investment Soars, But Governance Gaps Erode Billions in Potential ROI
• MIT
• More technology news