Raju Korti
Mathematics has always been less
about answers and more about the journey to reach them. For many of us, it has
been a strange companion, a mix of fear and fascination. Problems from calculus
books, whether from Wartikar brothers or Kaplan & Lewis, were never just
exercises. They were small battles. You struggled, you doubted, and then
suddenly, everything clicked. That quiet “click” was the real reward.
![]() |
| (Visual conceived by me) |
Because mathematics is not just about getting the right answer. It is about how you get there. Think of a simple situation. When you solve a tricky problem after hours of effort, there is a sense of ownership and pride. Now imagine typing the same problem into an AI tool and getting the answer instantly. It may feel efficient, but it also feels hollow. It is like watching a suspense film after someone has already told you who the murderer is. The story remains, but the thrill disappears.
The joy of mathematics rests on very human elements. It begins with curiosity, the urge to understand why something works the way it does. It grows through struggle, the mental effort that forces you to think deeper. And it peaks in that “aha” moment, when everything suddenly falls into place. If the struggle is removed, the “aha” loses its meaning. Without that journey, mathematics risks becoming a mechanical task rather than an intellectual adventure.
This concern is not limited to personal nostalgia. Leading mathematicians across the world have begun to voice similar worries through what is now known as the Leiden Declaration. Supported by global bodies like the International Mathematical Union, this declaration urges caution. It warns governments and institutions not to get carried away by the hype surrounding AI’s mathematical abilities.
The concern is rooted in the very foundations of mathematics. The discipline is built on trust. A proof is not merely a result but an explanation that can be checked, debated, and understood by others. AI systems, however, can produce answers that appear correct but may not always be reliable. They can blur authorship, skip proper credit, and make independent verification more difficult. In doing so, they threaten the core values of clarity, transparency, and accountability that mathematics depends on.
There is also a larger concern about control. Much of today’s AI development lies in the hands of private companies. The declaration cautions governments against blindly trusting these systems and stresses the need to prevent knowledge from being concentrated in a few hands. If research begins to depend heavily on such tools, even the direction of mathematics could change. Problems may be chosen not for their depth or importance, but because they are easier for machines to handle. This would mark a subtle yet profound shift in the discipline.
In response, the declaration offers a broad but clear direction. Governments must regulate the AI industry with care, ensuring transparency and reliability rather than accepting claims at face value. They should invest in public alternatives so that dependence on private technology is reduced. Academic standards must be protected, with AI-assisted work held to the same rigour as traditional research. At the same time, access to such tools must remain fair, so that they do not deepen inequalities among students and researchers.
Mathematics has survived many technological changes in the past. Calculators did not destroy it, nor did computers replace it. But Artificial Intelligence feels different because it touches the very act of thinking. It does not just assist the process; it risks taking it over.
At its core, mathematics has always been a deeply human endeavour. AI can certainly help us move faster and even open new doors of discovery. But if it takes away the struggle, the curiosity, and the joy of finding answers ourselves, then something essential is lost.
Solving a difficult problem from an old textbook may soon become optional. But the satisfaction of solving it on your own should never become obsolete.







