Master Machine Learning In Your Workplace (without Draining Your Human Batteries)

By Tamer Boyacı and Francis de Véricourt, ESMT Berlin

Let’s imagine that, for years now, Emily has led a successful German Mittelstand company, one of the so-called Hidden Champions that manufactures high-quality products for a niche customer base. A wave of AI disruption is sweeping across her industry, revolutionizing the way products are designed, manufactured, and delivered. To remain competitive, the company must adapt to this increasingly AI-driven landscape, but it fears tipping the delicate balance between embracing automation and preserving the craftsmanship and human touch that has won the company’s loyal following and retained its skilled workforce. Emily needs guidance on the strategies that will help her harness the power of human-and-machine integration while inspiring her workforce to welcome, not fear, this new digital age.

Despite the fear-driven media headlines, business leaders and workers can indeed win from a harmonious collaboration between humans and machines. This success, however, hinges on leaders identifying and capitalizing on the skills of both. Drawing on recently published research in the Journal of Management Science – “Human and machine: The impact of machine input on decision-making under cognitive limitations,” written together with Caner Canyakmaz, an associate professor of operations and management science at TBS Education – we share three evidence-based strategies that optimize AI-integrated decision-making processes and bolster human cognitive abilities. Using these strategies, leaders like Emily can better manage the challenges of AI integration while keeping the human expertise that drive their organization’s success.

Augmentation, not replacement

Recognize that AI should be seen as a tool to help human decision-making, not replace it. Highlight the power of human judgment and expertise while enjoying the value that machines and algorithms bring to the work.

While machines are great at calculations and analyzing large data sets, human judgment and expertise remain essential for nuanced decision-making in complex and dynamic business situations. In the medical sector, machines can excel at analyzing medical images and detecting abnormalities with high accuracy. For example, in 2018, a research team from France, Germany, and the UK published a study showing the successful training and output of an AI tasked with diagnosing melanoma. While the AI outperformed 58 dermatologists, the researchers saw the greatest benefit in integration, not displacement of skilled physicians. “We foresee AI being integrated into a routine consultation, with clinical examination, photography of suspicious lesions, and AI support to assist the clinician reach an appropriate management decision.” Doctors remain important because they consider not just the images and disease data but also the complex interaction of the patient’s family health history, current symptoms, living conditions, and other contextual details.

By encouraging a team approach that combines the unique strengths of humans and machines, organizations can improve decision-making. Yes, machines excel at specific tasks. But humans possess invaluable abilities such as flexibility, adaptability, emotional intelligence, and perhaps more importantly, an ability to properly frame problems. Leveraging these complementary skills results in more accurate and robust outcomes.

Optimization

Take advantage of machine-based decision-making when it can reduce human cognitive stress. Consider workload and mental constraints while tailoring machine integration to specific situations.

Using machine predictions in decision-making processes can be both good and bad. In diagnostic tasks, such as those frequently faced in medicine, our research shows that machine input can make decisions more accurate by reducing false negatives but can also lead to more false positives that place physicians under stress, resulting in slower processes, delays, and information overload. We find that these challenges are especially evident when time is limited or multitasking is required, as cognitive capacity is compromised. This applies not just in medicine, but also in fields like manufacturing, HR, and legal, where incorporation of machine learning and AI is increasingly prevalent.

To optimize machine integration, leaders should know the workload and cognitive limitations faced by their teams and offer alternatives, such as sufficient time for decision-making, minimizing multitasking, and making a supportive work environment. When time is tight, leaders should carefully evaluate how machines can improve human decision quality within the available timeframe. Similarly, where problems are complex and ambiguous, machine-learning algorithms can help analyze big data and uncover patterns that humans might overlook. Using machines thoughtfully can enhance teams and reduce worker stress.

Training and education

Equip your workforce with the knowledge and skills they need to work well with machines. Offer training programs that teach machine learning concepts, so workers can use machine-based predictions in their decision-making with confidence.

While our research does not specifically address it, a training and education strategy aligns with a broader understanding that empowering leaders (and their teams) through education is vital to organizational success. Understanding machine learning is crucial for effective human-machine collaboration in decision-making because it enables humans to better interpret machine-based predictions and integrate them into their decision-making processes. It also allows them to identify potential biases, errors, or limitations in these predictions. Additionally, a basic understanding of machine learning can foster trust in these systems, encourage adoption, and help individuals adapt to new workflows or processes enabled by these technologies.

Leaders should provide comprehensive training programs that cover the practical applications of machine learning in their industry alongside lessons on data privacy, ethics, and mitigating biases in AI systems. Well-designed executive education programs offer a multidisciplinary approach, drawing on lessons from management science, computer science, and decision science. Such a holistic perspective on the role and potential of AI in modern decision-making can help teams understand the utility and limitations of AI, encouraging trust in integrated, human-machine collaboration.

Conclusion

As AI disruption continues to sweep across industries, the delicate balance between embracing automation and preserving human craftsmanship becomes a critical concern for business leaders. However, evidence-based strategies derived from recent research can guide leaders in mastering machine learning in the workplace without draining their human batteries. By adopting a collaborative approach that emphasizes augmentation, not replacement, leaders can leverage the unique strengths of both humans and machines to enhance decision-making processes. Recognizing the importance of workload and cognitive constraints, leaders can tailor machine integration to specific situations, optimizing outcomes while reducing human cognitive stress. Equipping the workforce with the necessary knowledge and skills through comprehensive training programs fosters effective human-machine collaboration and instills trust in AI technologies. Moreover, establishing a culture of continuous evaluation and feedback enables leaders to monitor the impact of machine learning, address challenges promptly, and refine integration strategies over time. By implementing these strategies, leaders can navigate the evolving landscape of AI integration, harness the power of human-and-machine collaboration, and drive their organizations to thrive in this new digital age.