The Right Shift of Merit: The Lost Bell Curve

Dear Friends,

This is the season of results. From as early as 5th standard, both students and parents are preoccupied—not about passing or failing, nor about third, second, or even first class—but about the race for 90%+ and 9.0+ CGPAs.

I still remember my own 10th-grade results. In those days, the results came in a special newspaper edition. Pages filled with roll numbers, neatly categorized: third class, second class, and first class. I’m not sure if it was a confidence issue or a competence issue, but we would start by nervously checking the third-class list, then second, and finally, with bated breath, the first-class section. But the most suspenseful category of all was the "WITHHELD" section—a kind of academic Trishanku Swarga. Neither passed nor failed. Caught between suspicion of malpractice or a technical error. Pure suspense!

But oh, the joy of simply passing back then! After my intermediate (Plus 2), my father sent a telegram with just two words: First Class. I was on vacation in Chhattisgarh with no phone, no easy communication, and no reserved travel. My cousin helped me catch a ride on a steam engine (literally sitting next to the driver) to Raipur, then I took a passenger train to reach Vizianagaram. For that entire journey, I was basking in those two words: First Class. I also owe a small thanks to my classmate, Dr. Ravi Sankar (now Professor at IIT Madras), whose roll number was next to mine. Both names are alike. His math help made a real difference.
Where is that kind of joy now?

A few days ago, I came across a post by Dr. Madhuri Parti. She writes:

“Students scoring 95%, even 96%, are anxious, dissatisfied, and in some cases — battling depression. What changed? In 2024 alone, over 2.4 lakh Class 10 students scored above 90%. Over 2.2 lakh Class 12 students crossed that same mark. On paper, it looks like a nation of academic brilliance. But the reality is far more complex — and troubling. When 90% becomes average, we erode the value of genuine understanding. The dignity of skilled work. The joy of learning for life, not just exams.”

How true.

As teachers, we’re encouraged to follow the bell curve model while grading. It’s designed to standardize results across large classes, and to prevent grade inflation or deflation. It assumes that a normal distribution exists—and tries to fit everyone into it.
But reality rarely conforms to theory. When most of the class scores above 90%, trying to fit grades into a bell curve can feel unfair. High-performing students may still get pushed down the grade ladder—not because of their competence, but because of the statistical mold we’re forced to apply. No one feels good in the end.

Initially, I believed this was an India-specific problem. But it’s not. China, the U.S., and even European systems face the same challenge. In machine learning terms, it’s like an overfitted model—excellent results on the training data, but poor generalization to real-world scenarios.

We’re creating excellent academic pathways for higher education, but failing to build flexible, skill-based routes that cater to the socio-physical-psychological uniqueness of individuals.
And then there's the meme making rounds:

“This generation’s marks are like: 99.4, 99.5, 99.8, 99.9.
In our times, we only used to get fever like this.”

Maybe it's time we asked ourselves: Are we chasing the right kind of merit?

Ravi Saripalle

The Numbers Didn’t Reveal the Truth—Empathetic Observation Did

Dear Friends,

Over the past two weeks, I had the opportunity to speak about Design Thinking with diverse groups—from enthusiastic school kids in Hyderabad to a dedicated faculty team at Gayatri Vidya Parishad. Today, I had the pleasure of addressing members of the CMA Association (The Institute of Cost Accountants of India, Visakhapatnam Chapter) on the topic “Design Thinking for CMAs.”

In fields like medicine or consumer products, explaining Design Thinking is relatively straightforward. It’s a human-centered approach to experiential innovation—where human centeredness means empathy, and experiential innovation is the confluence of desirability, affordability, and technical feasibility.

But how do we explain Design Thinking in the context of Cost & Management Accountancy? I shared two stories to illustrate. Here’s the first:

Story 1:
A leading pharmaceutical distributor in Andhra Pradesh experienced a steady decline in net margins—despite increasing sales volumes. The finance team dove deep into the P&L, comparing year-over-year costs, vendor discounts, and pricing models. They adjusted overhead allocations, renegotiated logistics contracts, and even restructured sales incentives. But the problem persisted.

Enter Sudha (a fictional character), the company’s CMA—and more importantly, a Design Thinker at heart. She knew the numbers only told part of the story.

Sudha decided to look beyond the reports. She visited major pharmacies the company supplied to. She spoke to doctors—both those prescribing the products and those who weren’t. She interviewed patients about their experiences. She listened to store managers about ordering delays and invoicing issues. She even shadowed the warehouse dispatch team for a day.
What she uncovered surprised everyone.

Sales reps were being incentivized to push slow-moving, high-margin drugs to meet monthly targets. This resulted in high return volumes from pharmacies, which drove up reverse logistics costs. The manual reconciliation process for returned stock led to delays, missed credits, and write-offs. Pharmacies began favoring rival distributors who offered more flexible, responsive systems.

None of this had shown up in the reports.

The real issue wasn’t pricing—it was a broken user experience across the supply chain.

While AI could detect anomalies in the P&L, it was Sudha’s empathetic observations, interviews, and field visits that revealed the underlying problems. That’s the beginning of the Design Thinking approach.

I’ll share the second story in another letter—but it focuses on the idea of upholding trust.

Today, fraud control isn’t just about detection—it’s about designing trust.

When you’re a design thinker or designing a banking product, you face four possible outcomes:

1. You correctly flag a fraudulent transaction (your duty as a banker).
2. You wrongly flag a legitimate transaction as fraud (a failure of empathy—imagine a customer in a hospital being denied a transaction for surgery, or a traveler stranded because they couldn’t book a ticket).
3. You correctly allow a legitimate transaction (also your duty).
4. You wrongly allow a fraudulent transaction (potentially enabling illegal activities).

In short, while AI automates detection, design thinkers elevate prevention—by designing systems grounded in empathy, trust, and user understanding.

Let’s design that trust.

Ravi Saripalle

Why Bad News Hits Harder: The Science of Negativity Bias

Dear Friends,

Once, while walking through a narrow street lined with closely packed houses, I came across a typical scene: a heated argument between a husband and wife. Their voices echoed across the neighborhood. At least fifteen people had gathered to mediate. Almost every neighbor within earshot had stepped out of their homes to watch.

This isn't an isolated incident. You’ll find similar commotion near community water taps, in middle-class apartment car parks, at garbage collection points of independent houses, or even in affluent neighborhoods, where disputes over a few feet of land can turn ugly. The common thread? Negative talk draws attention.

Now pause and reflect—how often do neighbors genuinely celebrate your achievements? Perhaps occasionally, out of social courtesy. But more often than not, subtle envy overshadows genuine joy. It’s not just them. No one is immune to this bias. As uncomfortable as it sounds, people are instinctively more drawn to bad news.           

The Data Doesn’t Lie

A 2022 study analyzing 23 million headlines from 2000 to 2019 found a staggering 314% increase in negative sentiment. Headlines containing anger increased by 104%, fear by 150%, and sadness by 54%.

Even more striking, each additional negative word in a headline was found to increase its click-through rate by approximately 2.3%.

In one experiment, a Russian news outlet published only positive stories for a day. The result? A 66% drop in readership—a clear indicator of public preference for negativity (BBC.com, Dec 4, 2014).

The Psychology Behind It

Why do negative statements capture our attention more powerfully than positive ones?

The answer lies in a well-established psychological concept called negativity bias. Evolution hardwired our brains to prioritize threats—essential for survival. Negative events trigger faster and stronger responses in the amygdala, the brain’s emotional alarm system. As a result, negative information stands out more and spreads faster than its positive counterpart.

When we began the Inspire to Innovate Storytelling Movement (http://i2iTM.blogspot.com) in 2011, we were committed to sharing only positive stories. But we struggled—first to find such stories, and then to sustain them. Why? Because staying in a consistently positive vibration is difficult. We’re all human, and we’re all wired the same way.

Only a few spiritually elevated individuals seem to transcend this bias. I’ve seen people who diligently follow the Bhagavad Gita or their respective spiritual paths—living with purpose, contentment, and inner peace. These are exceptions, but they offer us hope.

On a lighter note, did you come to this point of reading because of today’s story title which has a negative word:)-

Ravi Saripalle

Already Scary—Did They Have to Call It an "Agent"?

Dear Friends,

Recently, I was talking to a well-educated professional. We were discussing how AI is going to impact jobs. I told him that AI agents will soon start doing many simple, everyday tasks that humans do today. Immediately, he asked me, “What qualification is needed to become an AI agent?” He also asked, “What will happen to real estate agents, tourist agents, insurance agents, freight forwarding agents, etc.?”

He is quite educated, but still confused. So, I thought this week’s story should explain this clearly.

First of all, AI Agent is not a human trained in AI!
It is a software program that works in the background, but does the same job as a human agent.

Recently, Puneet Agarwal, Director at Amazon, wrote an article explaining this beautifully. I am summarizing it here using simple examples:

1. Basic Human Helper = Old Software
You tell your helper: “Call Udipi Hotel at 7:30 PM.”
If it’s closed, he just comes back and asks, “What next?”
This is how old software works — it only follows exact steps. No thinking or adjusting.

2. Smart Friend = AI That Understands (LLM - like ChatGPT)
You say: “I’d love to eat Ravva Dosa.”
Your smart friend knows your taste and suggests a great restaurant — but you have to book the table.

This is like ChatGPT — it understands what you want, but it can’t take action.

3. Smart Friend + Assistant = AI + Agent
You say: “Book dinner for me.”
Your smart friend asks a helper to do it. The helper can act, but he needs clear instructions on how to book at that restaurant.

This is AI + Agent. The AI understands you, and the agent does the work — but only if you teach it each step.

4. Smart Friend + Assistant + Guidebooks = AI + Agent + MCP
Now your helper has official guidebooks from restaurants, airlines, etc., which explain exactly how to book, check flights, reschedule, etc.
He does everything without asking you again.

This is the future — AI + Agent + MCP (Model Context Protocol). MCP is like a digital manual that tells the AI what to do, step by step, without extra code.

Now think — what will happen to real estate agents, insurance agents, or freight forwarders? The same AI agent can replace their routine tasks too.

That’s why I tell my MBA students:
“Learn to build AI agents. You already have domain knowledge. If not, engineers will take your job.”

And I tell my engineering students:
“Learn business knowledge and build AI agents — or MBAs with AI skills will take your job.”

Of course, it is also applicable to a medical student, a Chartered Accountant, a biochemist, or any other student.

Universities must change, too. It’s not just about multi-disciplinary anymore. Now we need education that mixes skills across fields (interdisciplinary) and even beyond fields (transdisciplinary).

On a lighter note, the first James Bond novel came out in 1953. It was Casino Royale, introducing Agent 007. But today’s AI agents might be even more thrilling — and definitely more dangerous to jobs!

Agents are scary, right?

Now think — how long before an AI + Agent + MCP replaces you?

– Ravi Saripalle