From
Send to

[Lim Woong] AI and keeping humans in the loop

Nov. 12, 2024 - 05:30 By Korea Herald

I teach graduate courses about AI and mathematics education, and my students -- mostly working professionals -- often express concerns about their future, not the distant future but the next two or three years: "Is AI going to take over my job?" Especially when some AI tools already seem to outperform their best work, it is quite a valid worry. My response is that while AI might make some jobs obsolete, it will also create new opportunities, such as AI device designers, AI ethics specialists or prompt engineers. However, these job titles still sound exotic to most of my students, and what keeps coming up in our discussions is: How do we keep humans in the loop when AI technologies are developed and implemented?

In this age of artificial intelligence, the partnership between humans and AI is a fast-changing industry but still remains a relatively new idea. From education and health care to the creative arts, AI is reshaping how we tackle challenges, streamline operations and generate random new ideas. But as we dive into this tech-driven world, we can't forget the key role of human insight and oversight.

David De Cremer and Garry Kasparov have written about the concept of Augmented Intelligence in Harvard Business Review, which they called the third type of AI, combining Artificial Intelligence with Authentic Intelligence -- human intelligence. The partnership allows AI to reach its true potential, not by replacing human input but by incorporating human feedback as part of its design and functions. Authentic Intelligence could provide the empathy, ethical considerations and creativity that make AI more effective so that human values guide AI's development and application.

Take classroom teachers, for example. Human teachers play an important role in AI-integrated learning environments. While AI can assist in personalized instruction, grading and providing instant feedback, it is teachers who offer emotional support, mentorship and a sense of belonging. It is our own very human teachers who make sure AI textbooks and tools are developed and used in ways that nurture curiosity and reasoning, helping students gain knowledge, engage in thinking and build relationships with peers and people in the community. Teachers may use the analysis provided by AI within the learning management system, but it remains their professional responsibility and discretion to adapt teaching strategies to individual student needs and impact students learning in ways that AI cannot fully replicate.

In health care, AI is making huge strides in diagnostics and patient care. Algorithms analyze imaging data with impressive accuracy, often picking up on details that even the best-trained doctors might miss. AI-driven robots assist in surgeries, reducing the chances of human error. But the value of human doctors can't be overstated: the empathy they bring, the nuanced decision-making in complex cases and the trust they build with patients are irreplaceable. That said, health care professionals need to shift from just being expert diagnosticians and skilled surgeons to being interpreters and advocates for patient-centered AI.

The creative arts are another area where human-AI collaboration becomes relevant. AI can generate music, paintings, and even some poignant poems, often with stunning results. But art is about more than technical mastery; it is an expression of human experience and emotion. When artists collaborate with AI, they can push creative boundaries, but it is their vision and emotional sense of humans and nature that bring true meaning to the work. AI can help create, but it is human artists who infuse that creation with a personal story and purpose that resonates with his/her patrons and the public on a human level.

At the end of the day, AI systems are built to help humans, and their value lies not just in efficiency but also in human values and agency. Ge Wang, a professor of music and computer science at Stanford University, highlights the benefits of human involvement in AI. First, it means significant gains in transparency. Each step that incorporates human interaction demands that the system be designed to be understood by humans to take the next action. Next, incorporating human intelligence and interaction with AI ameliorates the pressure to build “perfect” algorithms. Since the system relies on human guidance, it only needs to make meaningful progress to the next interaction point. Ultimately, humans and AI undertake complex or creative tasks together, making it harder for the process to remain a mysterious black box.

Finally, keeping humans in the loop isn't just about managing risks; it is about making AI work in ways that genuinely maximize human potential. AI can analyze and predict, but humans bring context, sentiment, morality and a sense of purpose and self. As we integrate more AI into our lives, I want my students and their colleagues to understand that the real power of AI lies in using it as a close partner -- not as a replacement. Indeed, this collaboration will let us harness AI’s transformative potential while holding onto your humanity, dignity and, most importantly, our jobs.

Lim Woong

Lim Woong is a professor of the Graduate School of Education at Yonsei University in Seoul. The views expressed here are the writer’s own. -- Ed.