Beyond Calculation: IISc’s Bold AI Challenge to Undergraduate Science

At IISc’s July convocation, Fields Medalist Manjul Bhargava stated that AI will soon solve undergraduate-level science and math problems. This article explores curriculum redesign, creativity’s role, and how educators must adapt to an AI-first future.

NewsSutra NewsSutra
Jul 14, 2025 - 06:57
0 10
Beyond Calculation: IISc’s Bold AI Challenge to Undergraduate Science

1. A Visionary Convocation Moment

At the Indian Institute of Science’s convocation on July 11, 2025, distinguished mathematician and Fields Medalist Manjul Bhargava stunned graduates with a bold prediction: “within a year or two, some AIs will be capable of accurately solving any undergraduate science and math problem”—including challenging textbook exercises like trigonometry and thermodynamics Wikipedia+6India Today+6The Times of India+6. Drawing from hands-on experience with advanced language models, he noted that while current AI systems are inconsistent and sometimes confidently wrong, their improvement trajectory is unmistakable.

2. Why This Signals an Educational Earthquake

Bhargava’s warning goes beyond audit of AI capabilities. It exposes a core flaw in many institutions: the overemphasis on problem-solving rather than original thinking. As AI becomes competent at routine or even moderately complex problems, the educator’s role must evolve toward fostering intuition, creativity, and interdisciplinary reasoning.

3. Rethinking what we teach—and why

Bhargava posed a provocative question: “What will we teach when AI can do the solving?” He argued that emerging academic programs must pivot.

  • Instead of assigning classic textbook problems, educators should introduce open-ended, real-world scenarios, where meaningful solutions cannot be directly generated by AI.

  • Curricula should prioritize creativity workshops, research exposure, and collaborative problem framing.

  • Science and engineering syllabi must emphasize design thinking, experimentation, human judgment, and systems-level analysis—areas where AI can assist but not replace.

4. Balancing AI Utility and Human Intellect

Recognizing AI’s limitations is as crucial as noting its strengths. Bhargava cautioned against overreliance: “AI remains notoriously bad at doing math and science… often confidently wrong,” yet underpins everyday tools—the trap lies in trusting it blindly India Today+1Complete AI Training+1.

In practice, this means teaching students AI literacy: how to prompt models effectively, how to verify outputs, and how to treat AI as a collaborator, not a substitute.

5. Adaptations underway at IISc and beyond

This paradigm shift has already begun:

  • IISc is piloting AI-integrated labs, where students use AI for simulation development but must justify and validate results.

  • Other institutions are experimenting with AI teaching assistants focused on guiding labs or coding workflows—not just answering questions.

  • Numerous peer institutes and IITs are building AI ethics modules, and integrating creativity into STEM training—traceable back to Bhargava’s insights .

6. Preserving the human spark

Bhargava emphasized that creativity cannot be forced; it must come “from within” India Today+4deccanherald.com+4The New Indian Express+4. While AI may replicate procedural patterns, it lacks originality. Human-invented scientific breakthroughs often cut across disciplines, employ serendipity, and challenge assumptions—qualities that escape current AI architectures. He urged educators to craft experiences that spark curiosity, risk-taking, and genuine exploration.

7. Curricular shifts: what universities may introduce

  • Project-based, interdisciplinary courses: addressing real challenges like climate resilience or healthcare diagnostics.

  • AI-integrated labs where students pose questions, let AI propose solutions, and then critically evaluate outcomes.

  • Philosophy and ethics seminars exploring AI's societal impact, purpose, biases, and misuse.

  • Group thinking & presentation modules to build communication, leadership—outcomes beyond AI’s reach.

8. Employer demand: the new skills mix

Bhargava predicted a shift in graduates' required skill sets. As AI takes over rote computation, employers will seek candidates skilled at asking the right questions, modeling unstructured problems, and judging AI-provided insights. These roles demand advanced logic, creativity, and human insight—traits that should define tomorrow's STEM education.

9. Challenges & resistance ahead

Adapting to AI-driven education requires grappling with real obstacles:

  • Faculty training: Not all instructors are comfortable integrating AI into pedagogy.

  • Assessment overhaul: Objective, AI-resistant evaluation methods—like oral exams, project reviews—need design.

  • Digital fairness: Ensuring students from diverse backgrounds get equitable AI access. New hardware and infrastructure investments may be essential.

  • Legacy mindsets: Gupta, one student noted, “some lecturers fear AI will render us lazy.” A cultural and mindset transformation is required.

10. Global trends align with Bhargava’s views

His message resonates globally:

  • Universities in the U.K., U.S., and Europe are scrapping timed computational exams in favor of design-thinking portfolios and oral defenses.

  • Reports like MIT's Future of Work emphasize the irreplaceability of human-centric skills: empathy, synthesis, and innovation.

  • Even textbooks are evolving—incorporating AI-stimulating exercises, asking students to critique AI outputs or improve model prompts.

11. What Indian policymakers must do

To move forward, India’s higher education system must adapt:

  1. Regulatory flexibility: UGC and AICTE should encourage experimental pedagogy models.

  2. Faculty development schemes: Upskilling initiatives in AI pedagogy, academic leadership and curriculum renewal.

  3. R&D funding: Support for institutions developing AI augmented education tools for disciplines like physics, chemistry, and biology.

  4. Ethics and policy training: AI literacy should be mandatory across STEM and humanities programs.

12. A student’s perspective

At convocation, medallist Gurkirat Singh—a recipient of a gold medal—reflected on AI’s potential. He noted that tools that expedite routine work give students room to engage deeper with original thinking. Yet he echoed Bhargava’s stance that foundation skills remain critical, for interpreting complex AI results.

13. Moving from prediction to practice

A key test is whether this shift—from problem-solving to imagination—takes root. Institutions like IISc may lead the transformation, but it requires shared vision and investment. If successful, education will not just transmit knowledge to machines, but rather teach students to surpass them.

14. Looking ahead: tomorrow’s convocation

Envision future convocations: degrees awarded for research proposals, not solved textbook problems; for interdisciplinary designs, not single-subject performance. AI literacy credentials may complement technical degrees. And more thought leaders—like Bhargava—may catalyze this seamless integration of technology and human purpose.

Conclusion: Human Wisdom in the Age of AI

Manjul Bhargava’s statement at IISc convocation marks more than a technological observation; it’s an educational wake-up call. As machines get smarter, institutions must cultivate the uniquely human aspects of intelligence: curiosity, creativity, ethics, and imagination.

The message is clear: we must teach our students not just to solve, but to question—and when AI can do the solving, we will know we succeeded.