The Catastrophic Decline in C-L-C Thinking (Critical, Logical, and Computational) Among Students: Is It a Cause for Worry in India?

Dear Friends,

In a study of approximately 1,000 high school students, Stanford University's Bastani et al. (PNAS, 2025) found that AI assistance improved practice performance by 48%, but students using standard AI scored 17% lower on subsequent exams without AI, highlighting the risk that AI dependence can weaken independent problem-solving.

Gartner warns about "cognitive surrender": accepting AI outputs at face value without evaluating them yourself.

PwC is investing more than $100 million in a new critical-thinking certification program for its 85,000+ US employees, designed to help them question AI-generated conclusions, spot errors, and make better decisions. The certification logic goes like this: For example, a tax professional receives an AI-generated analysis of a complex tax position. The recommendation looks convincing, but some of the supporting information may be incomplete or need verification. Before reaching a conclusion, the participant would question the sources, consider alternative interpretations, and work out what additional information is needed.

The Next Competitive Advantage May Be Human Judgment. The next phase of workforce transformation will not be defined by how quickly employees learn to use AI, but by how effectively they learn to question it, interpret its outputs, and take responsibility for the decisions that follow.

What does it mean for Computer Science/Engineering students? AI generates an authentication system in minutes. The code might work and pass basic tests. But who checks the security risks, missing edge cases, or incorrect assumptions?

In recent times, we have seen four types of students:

• Type 1 – Students who learn on their own.
• Type 2 – Students who are learning to program using prompts. AI generates the output, but they learn by decoding the generated output.
• Type 3 – Students who believe in the generated code and use it without questioning.
• Type 4 – Students who ignore coding, as they believe AI will take over anyway.

Today, types 2, 3, and 4 account for more than 80–90% of students. This is the catastrophe. What is the consequence? Loss of Critical Thinking, Logical Thinking, and Computational Thinking (C-L-C Thinking). If India does not address this, the consequences will be worse than those of a major natural disaster. So, when should we address this?

Undergraduate – Too late.

• Plus 1 and 2 – Focus has shifted towards memory-based national tests.
• Before Class 7 – Too early, as students need to focus on RAW (Read, Arithmetic, Write) foundational skills.
• Classes 9 and 10 – Desired group, but students are still in a transition state towards memory-based national tests.
• Class 8 – Ideal and best beginning.

Do you agree?

Project C-L-C Thinking is being cooked up on a pilot basis. I will keep posting on this.

Ravi Saripalle