The rapid proliferation of artificial intelligence has
turned what were once speculative questions into urgent, everyday
conversations. In boardrooms, classrooms, and chai stalls alike, we now ask: Has
India missed the foundational model bus? Has AI surpassed human intelligence?
And what happens to jobs as machines become smarter than us at most cognitive
tasks?
These are not idle musings. They are existential questions
for a country of 1.5 billion people, standing at the cusp of a technological
revolution. And few voices articulate this moment with as much clarity and
conviction as Vishal Sikka, former CEO of Infosys, veteran of SAP and Oracle,
and now founder of Vianai Systems. His recent podcast conversation offers a
treasure trove of insights, and I’d like to unpack and reflect on them here, in
the spirit of strategic curiosity and national urgency.
Has AI Surpassed Human Intelligence?
Let’s start with the provocative question: Has AI become
smarter than us?
Sikka’s answer is refreshingly grounded. He reminds us that
today's large language models (LLMs) are essentially “lookup machines”, brilliant
at pattern recognition, but devoid of true understanding. They can generate
answers, yes, but they don’t know anything. They lack grounding in the
physical world, in causality, in embodied experience. They are not sentient,
and they are not superintelligent.
What they are, however, is astonishingly efficient.
Consider this: our brain runs on about 20 watts of energy. Training GPT-5, by
contrast, consumes energy orders of magnitude higher. somewhere between 10¹² to
10¹⁸ times more. That’s a trillion-fold gap in efficiency. And yet, despite
this brute-force power, AI still stumbles on basic reasoning, context, and
nuance.
So no, AI hasn’t surpassed human intelligence. But it has
become a powerful tool, like a calculator, then Excel, then Google, and now
this. The question isn’t whether it’s divine. The question is: What can we
build with it?
Has India Lost the Foundational Model Story?
This is where the conversation gets interesting and
controversial/.
Sikka is unequivocal: India must build its own
foundational models. Not just because we can, but because we must. To be a
passive consumer of AI built elsewhere is to surrender our agency in shaping
the future. And India, he argues, is large enough, deep enough, and important
enough to do it all. build the models, build the applications, and build the
services.
We have unique advantages:
- India
Stack: A digital infrastructure unmatched globally, offering rich,
structured data.
- Linguistic
diversity: Hundreds of languages and dialects, ripe for training
multilingual models.
- Cultural
archives: Manuscripts, documents, and oral traditions that no other
country possesses.
And yet, the expertise to build frontier models is
shockingly concentrated. According to Sebastian Thrun, only about 3,000 people
globally can build such models, and 80% of them are in San Francisco and
London. This is not just a talent gap. It’s a geopolitical vulnerability.
India must democratize this capability. Stanford teaches a
course on building foundational models. Why shouldn’t IITs, IIITs, and NITs do
the same? Why shouldn’t we have open-source frameworks, indigenous datasets,
and public-private partnerships to accelerate this journey?
The Future of Jobs: Catastrophe or Opportunity?
This is perhaps the most emotionally charged part of the
conversation. Sikka doesn’t mince words: Mass unemployment is a real and
imminent risk. And paradoxically, it’s the educated class, those trained
for certificate-based jobs like database administration or network maintenance,
that are most vulnerable.
But here’s the twist: AI could empower artisans more than
engineers.
Imagine a village woodworker using AI to design, market, and
sell his craft globally. Imagine a weaver translating her product descriptions
into 20 languages. Imagine a painter understanding global trends and adapting
her style. These are not fantasies. These are real, empowering use cases.
The challenge, then, is not just technological. It’s
societal. We must shift from training people to “make a living” to training
them to “make a life.” That means teaching them how to use AI to augment their
creativity, productivity, and agency, not just to pass certification exams.
And yes, while many jobs will become irrelevant, many new
ones will emerge. Transitioning legacy systems and reimagining business processes for AI enablement
of existing enterprises, managing AI
ethics, curating datasets, fine-tuning models, these are all new frontiers.
Services companies will play a pivotal role in this transformation, but they
must evolve from body-shopping to capability-building.
What Should India Do Next?
Let me offer a strategic synthesis, drawing from Sikka’s
wisdom and some of my reflections:
- Invest
in foundational models: Not just one, but many. Across languages,
domains, and modalities.
- Democratize
AI education: From elite labs to vocational centers. Teach people how
to build on and around it.
- Empower
the informal sector: Use AI to elevate artisans, farmers, and
micro-entrepreneurs.
- Reimagine
job training: Move from certificate-based skills to capability-based
learning.
- Build
public infrastructure for AI: Open datasets, ethical frameworks, and
compute access must be national priorities.
Final Thought: The Building Is Not Smarter Than Us
Sikka ends with a beautiful metaphor: the building we’re
sitting in is more powerful than us. But we don’t worship it. We use it. We
live in it. We shape it.
AI is the same. It’s not God. It’s not superintelligence.
It’s a tool. And like all tools, its value lies in what we do with it.
India’s AI moment is here. Let’s not squander it. Let’s
build, with clarity, courage, and conviction.
“The true measure of intelligence is not in what we can
automate, but in what we choose to preserve.” Anonymous
“AI hasn’t taken over the world yet—but it has taken over my
browser tabs, my inbox, and my sleep.”Anonymous
Sources:
PM
Modi and Vishal Sikka Chat About India's Bright AI Future
Vishal
Sikka’s Advice for India on Foundational Models
A
Visionary Meeting: Vishal Sikka & PM Modi Discuss AI's Future
