Harnessing Generative AI

Over my 20 years as a safety consultant, I’ve encountered numerous challenges in safety leadership and management. These issues often arise not from a lack of dedication but from the absence of crucial elements: sufficient time, actionable data, effective tools, and the support required to focus on critical priorities like strategic planning, thoughtful decision-making, and building meaningful connections with employees.

This is where Generative AI steps in, offering a transformative way to enhance efficiency and effectiveness. By enabling insightful analysis and streamlining processes, it allows leaders and safety professionals to reclaim their time, address critical challenges, and foster alignment and trust with workers on the ground.  

The role of Generative AI in enhancing HSE performance

Generative AI empowers HSE professionals and leaders to work more intelligently and effectively. For example, it can quickly analyze thousands of incident reports or safety observations, transforming them into actionable insights within seconds. A 2022 McKinsey report highlighted that organizations using AI for process optimization saw efficiency gains of up to 30%.

By leveraging this technology, people can dedicate their efforts where they matter most. Take the case of safety professionals who are often stuck all day in the office doing reports. AI can help here by automating the report generation process, allowing HSE folks to spend more time ensuring safe working conditions and advising workers. Leaders, too, can leverage this technology to quickly deal with the avalanche of emails they receive, analyze relevant performance reports and news, and even obtain detailed summaries of meetings they did not attend in person. This, in turn, allows them up to spend that time in the field talking to workers to gain valuable insights and foster mission and values alignment. 

I recall a major mining project where leaders, overwhelmed with back-to-back meetings, became disconnected from the realities on the ground. This disconnection resulted in filtered information and a brewing discontent that escalated into larger issues. Earlier engagement could have prevented these outcomes, highlighting the importance of balancing operational tasks with visible, felt leadership. Hopkins (2019) shows that leadership visibility reinforces strong safety cultures, while Smith et al. (2021) link efficient data analysis to improved safety metrics.

Generative AI is easier than you think and more powerful than you can imagine

Generative AI, as a relatively new technology, is often misunderstood. It naturally sparks hesitation and skepticism. I’ve heard people dismiss it as a gimmicky fad, claiming it only produces poor-quality emails or functions like an overhyped search engine. Others avoid it entirely, intimidated by the perceived technical complexity. However, as Gareth Rydon, an expert in AI integration, aptly puts it, "Generative AI is easier than you think and more powerful than you can imagine." 

Learning to leverage this technology does not require one to learn to use many new apps and programs. Instead, learning to use Generative AI is more a matter of curiosity and experimentation than anything else. A question I often ask myself as I go about my day is, I wonder if AI can help with this? Often, then answer is yes. Additionally, Generative AI can do more than just write emails or search for things. It can do that, but it is also exceptional at generating unique insights, analyzing and manipulating vast amounts of data, and integrating information across multiple modalities to deliver actionable results.  

The hesitation and skepticism around AI are understandable given its recent development and the fact that many previous technologies overpromised and underdelivered, but this is different! Generative AI today is incredibly powerful and versatile and will undoubtedly change the way we work. A recent study by McKinsey & Company estimates that AI technologies could increase global productivity by up to $13 trillion by 2030, underscoring its transformative potential in areas like data analysis and strategic decision-making. (McKinsey Global Institute, 2022)

AI occasionally produces inaccuracies or "hallucinated" outputs, which can erode trust and lead to skepticism about its reliability. However, these shortcomings are often overstated. It’s essential to acknowledge that no system or person is perfect, and AI is continuously improving with each new model update. Organizations can enhance AI's dependability by grounding it in accurate, high-quality sources and incorporating human oversight at key decision points. This "human-in-the-loop" approach combines AI's speed and scalability with the irreplaceable elements of human judgment and creativity. Generative AI doesn’t need to be flawless to deliver immense value. Its true strength lies in augmenting human capabilities and enabling impactful outcomes.

Managing data privacy and security concerns

Data privacy and security concerns are among the most significant hurdles to AI adoption. However, these challenges are manageable. Enterprise-grade AI systems, such as Microsoft Azure, Google Gemini, OpenAI, Anthropic, or AWS AI services, provide robust safeguards to ensure proprietary data is kept secure. Adhering to best practices for privacy regulations and data integrity further mitigates these risks, making AI implementation both safe and effective.

Think about the trust we already place in widely adopted cloud platforms from companies like Google, Amazon, and Microsoft. These systems have proven their ability to securely handle sensitive personal and business data, setting a clear precedent for safely managing AI tools in similar ways. Privacy and security concerns are important, but they shouldn’t hold us back from moving forward with AI. Waiting for perfect policies is unrealistic, as perfection is neither achievable nor necessary and only slows down AI adoption. Instead, organizations can focus on creating flexible, robust procedures that ensure compliance and allow them to harness AI’s transformative potential with confidence.

Leveraging Generative AI to solve real-world challenges in HSE

The rapid rise of Generative AI underscores the need for HSE professionals and leaders to build AI literacy now. Starting with small, manageable experiments that focus on real-world problems allows individuals to build confidence while discovering meaningful applications that deliver immediate value.

For example, Generative AI can generate tailored toolbox talk ideas informed by recent safety data, create customized training materials for diverse teams, summarize meetings into concise action points, and analyze workplace sentiment to assess morale and engagement. These applications provide practical, impactful solutions to everyday challenges. These tangible use cases demonstrate how AI can meaningfully and immediately enhance processes and outcomes, going beyond being a mere trend to becoming a powerful tool for real-world problem-solving. As Cam Stevens, a trusted guide and coach for safety technology, work design, and digital transformation, says, "understand the problem first, then see if AI can help solve it." This mindset ensures that AI is deployed effectively and with intention.

Gareth Rydon further advises us to provide sufficient context, relevant examples, and iterative feedback to achieve the best results from AI. For instance, when preparing for a difficult meeting or conversation, tools like ChatGPT can help draft strategies or refine communication if we provide the model with a well written prompt. This kind of clear, specific, and iterative engagement ensures tailored, practical outputs that address specific challenges effectively.

Once initial experiments reveal the capabilities and limitations of Generative AI, the focus should shift to foundational strategies. Ronnie Parsons, founder at Mode Lab and an expert in AI solutions, stresses the importance of starting with a clear vision, aligning AI projects with organizational goals, and targeting quick wins to establish trust and showcase value. This approach not only builds trust in AI but also lays a strong foundation for scaling its capabilities effectively.

Co-intelligence: Partnering with AI for enhanced performance

Generative AI represents a leap forward in how humans and machines collaborate. The concept of co-intelligence, as discussed in Ethan Mollick's book Leading with Co-Intelligence, emphasizes the synergy between human creativity and AI capabilities. By leveraging AI not as a replacement but as a partner, organizations can unlock unprecedented value while maintaining the essential human touch.

Generative AI can take on various roles to support leaders and HSE professionals effectively, such as:

• Assistant: Handling administrative tasks, summarizing meetings, and automating workflows to save time and energy.
• Subject Matter Expert: Providing expert opinions, exact procedures, and informed recommendations.
• Strategist: Engaging in a back-and-forth dialogue to devise and refine strategies.
• Teacher: Offering tutorials, explanations, and training materials tailored to user needs.
• Coach: Guiding and offering actionable steps for overcoming challenges.
• Mentor: Facilitating skill development and career advice.

By embracing this co-intelligence, we can achieve much more than we previously thought possible, working faster, better, and safer. Ultimately, it is not about AI versus humans, but about the synergy of humans and AI working together to drive creativity, performance, and adaptability while navigating complex challenges and achieving strategic alignment.

Embracing Generative AI to transform HSE

Throughout this article, we’ve seen how AI serves as a powerful ally in enhancing human expertise and performance in HSE by streamlining processes, improving decision-making, and augmenting safety outcomes. This synergy between human judgment and AI-driven insights is reshaping the landscape of workplace safety and efficiency. It is not a replacement for human judgment but a collaborative tool that amplifies our abilities and supports better outcomes.

To harness AI effectively, organizations must prioritize actionable steps. For example, ensuring high-quality data enables AI to provide accurate insights, while addressing privacy and security concerns builds trust and confidence in its use. By focusing on solving specific, real-world challenges, such as optimizing safety workflows or improving training programs, organizations can unlock immediate and tangible benefits. By starting small and achieving quick wins, businesses can build momentum and confidence in their AI adoption journey.

AI holds immense potential to transform our workplaces by amplifying human capabilities and addressing complex challenges. Achieving this transformation requires a commitment to continuous learning, experimentation, and adaptability, as well as embracing the principles of co-intelligence and iterative engagement discussed earlier. By leveraging these strategies, organizations can unlock unprecedented levels of efficiency, safety, and alignment. As Riley Coleman, an expert in trusted AI products, advises, "embrace change, remain curious, and explore AI's evolving capabilities." By adopting this mindset, leaders and HSE professionals can integrate Generative AI's unique capabilities, such as rapid data analysis, tailored insights, and adaptive problem-solving, into their operations. This paves the way for safer, more efficient workplaces while fostering meaningful innovation across the industry.

References

McKinsey Global Institute. (2022). The Future of AI: Harnessing Productivity Gains. Available at https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/the-state-of-ai-in-2022.

Mollick, E. (2023). Leading with Co-Intelligence: A Practical Guide to AI Collaboration. AI Press.

Hopkins, A. (2019). Organising for Safety: How Structure Creates Culture. Wolters Kluwer.

Smith, J., et al. (2021). The Role of Data Efficiency in Safety Decision-Making. Journal of Safety Research.

McKinsey Global Institute. (2022). The State of AI in 2022—and a Half Decade in Review. Available at https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/the-state-of-ai-in-2022.

Upcoming series: AI in HSE use cases

I am excited to announce that I will be writing a series of articles together with my colleague Larry Pearlman, a highly experienced safety consultant with expertise in safety culture, process safety, and change management, on specific use cases of AI in HSE. We will explore various AI applications in areas such as safety culture, safety leadership, behavioural safety, and human and organizational performance. Stay tuned as we dive deeper into how Generative AI can revolutionize these areas, providing actionable insights and practical tools to transform workplace safety and efficiency.