In a knowledge-driven economy, innovation is paramount. Yet the popular conception of artificial intelligence (AI) remains largely transactional viweing it as a collection of tools designed for efficiency and automation. AI currently excels at providing answers and solving clearly defined problems.
But what if the true potential of AI lies not just in answering questions, but in asking them?
Imagine a different model of human–AI interaction: co-thinking. Rather than positioning humans as merely users and AI as tools, we could envision a collaborative partnership, where AI doesn't simply respond, but actively engages, prompts, challenges, and refines human thought processes.
Current AI systems, especially large language models (LLMs), have astonishing linguistic abilities and breadth of knowledge. Yet their predominant application is task-oriented, offering answers and completing user-defined jobs. This approach limits the cognitive potential of both human and machine.
In contrast, co-thinking would encourage AI to engage humans as intellectual partners. For instance, rather than passively awaiting commands, an AI trained for co-thinking would proactively ask thought-provoking questions, suggest alternative perspectives, and help humans explore new angles or unnoticed assumptions. Such an AI would not only enhance productivity but actively stimulate human creativity and critical thinking.
Singapore is well-positioned to pioneer such a shift. Educational institutions and business alike could deploy co-thinking AI models to facilitate deeper intellectual engagement, transforming workplaces and classrooms into dynamic spaces of inquiry and innovation.
Imagine university classrooms where AI co-thinkers pose challenging questions that encourage deeper student reflection and intellectual debate. Or professional meetings where AI actively participates in brainstorming sessions, pushing teams beyond surface-level ideas by asking insightful, even provocative, questions.
Realizing this vision requires shifting how we conceptualize and use AI systems. Instead of merely focusing on accuracy and responsiveness, training models should emphasize inquiry, curiosity, and intellectual humility. AI must learn to prioritize questions that enhance cognitive depth rather than offering straightforward answers.
This also implies developing a new skill for humans: learning to collaborate intellectually with machines. Humans must grow comfortable engaging with AI-generated queries, critically reflecting on their assumptions, and integrating novel perspectives raised by AI partners.
Such a model of co-thinking transforms the relationship from one of dependency or simple delegation to genuine intellectual collaboration. It holds the promise of not merely automating human work, but augmenting human cognition.
Let’s build human-centred futures—together.
Thanks for reading.
— HM