Many worlds of AI: For investors, the implications are significant

Two stories from the past few weeks capture something essential about where we are with AI.

The first concerns Salesforce, the enterprise software giant that aggressively embraced AI for customer service. CEO Marc Benioff proudly announced that AI deployment had allowed the company to cut support staff from 9,000 to roughly 5,000. Then reality intervened. Reports from late 2025 indicate that the company is now withdrawing from AI due to widespread failure.


The AI agents confidently gave wrong answers, dropped instructions when given more than eight steps, and lost focus when users asked unexpected questions. Customers complained that AI support took longer than the simple old search function. Salesforce is now retreating to rigid, rule-based scripting--essentially admitting they were, in their own words, "more confident" than the technology warranted.

The second story is a zeitgeist shift. Over the past couple of months, the conversation around AI and coding has transformed completely. People who were skeptical six months ago--senior developers who actually write code for a living--are now saying the age of human beings writing code is ending. Not in some distant future, but imminently. Entire features are being shipped by AI with minimal human intervention. The productivity gains are no longer incremental; they're structural.

How can both be true? How can AI fail comprehensively in customer service--seemingly straightforward--while revolutionising software development, which appears far more complex?

The answer is that we've been thinking about AI wrong. We treat it as a single phenomenon that will sweep through the economy at roughly the same pace. However, AI in business is not a single story. It's many parallel stories, moving at wildly different speeds. And the distinction has almost nothing to do with how intelligent the AI is.

I've written about this tension before. A year ago, I argued that "the fact that a revolution is real doesn't mean that every business claiming to be part of it will succeed." More recently, I observed that "the gap between what AI demos well in controlled environments and what it actually delivers when confronting the messy real world remains enormous." I now think there's a more precise way to understand this gap. It's not random. It's structural.

Consider what makes coding fertile ground for AI. Code is formally structured and machine-verifiable--it runs and passes tests, or it doesn't. The feedback loop is immediate. When AI makes a mistake, a developer (or another AI agent) notices, fixes it, and moves on. Errors are private and reversible. Now consider customer service. Customers don't speak in data schemas. Emotion, sarcasm, and cultural context matter enormously. One wrong answer can escalate to social media outrage or regulatory complaints. The failures are public and often irreversible.

The difference isn't intelligence. It's what I'd call error economics. AI thrives where mistakes are cheap, private, and correctable. It struggles where mistakes are expensive, public, and permanent.

We received a clear illustration of executive disconnect just a few days ago. During Bajaj Finance's Q3 call, CEO Rajeev Jain announced that AI had listened to 2 crore calls and generated 100,000 new customer offers. "We'll be able to listen to 100 million calls next year," he said proudly. The response on social media was predictable hilarity. As the entire country, except apparently Mr Jain knows, Bajaj Finance's incessant spam calls are the butt of countless jokes. Here was a CEO using sophisticated technology to optimize something customers actively despise. Machine learning works perfectly; the learning about customers is absent.

For investors, the implications are significant. When you hear "AI" attached to a business function, ask: what happens when it's wrong? If the answer involves customers, regulators, or reputations, progress will be slower than vendor PPTs claim. If the answer is "someone notices and fixes it," that's a different world entirely.

The story of AI in business is not one of universal acceleration. It's one of the selective escape velocities. Coding has left the atmosphere and gone into orbit. Customer service is still fighting gravity. Most other functions lie somewhere in between--mistakenly assumed to be closer to the rocket than they really are. The many worlds of AI are not converging. They're diverging. And that divergence will determine which investments succeed and which disappoint.

(Dhirendra Kumar is Founder and CEO of Value Research)