A few days ago, in San Francisco, I listened to early-stage founders compress years of work into three-minute pitches.
I was judging startups at STEP San Francisco. Artificial Intelligence (AI) was, unsurprisingly, everywhere. Yet after several days of conversations with founders and investors, the question I brought home to London was not really about AI.
It was about recognition.
Investors like to describe what they do as pattern recognition. Experience supposedly teaches us what a promising founder, market or company looks like.
There is an obvious problem with this. The better we become at recognising successful patterns from the past, the easier it becomes to overlook something that does not yet resemble them.
Call it pattern blindness.
It may be becoming one of the more important problems in technology.
For much of the software era, the ability to build was itself scarce. Turning an idea into a product required engineers, time, infrastructure and capital.
AI is beginning to reduce some of those barriers.
Small teams can prototype faster. Software development is becoming cheaper. Sophisticated capabilities that once required considerable technical resources can increasingly be accessed through existing models and tools.
That is good news for entrepreneurs. But it changes the nature of competitive advantage.
{{/usCountry}}That is good news for entrepreneurs. But it changes the nature of competitive advantage.
{{/usCountry}}When more people can build, knowing what to build becomes more valuable.
And when thousands of founders have access to similar underlying technologies, the advantage may increasingly belong to the person who understands something about the world that everyone else has overlooked.
That thought has particular relevance for India.
For years, India's technology conversation has contained a slightly defensive question: Can Bengaluru become another Silicon Valley?
Perhaps it should stop asking that.
India has something Silicon Valley cannot manufacture: Proximity to a vast number of difficult problems at extraordinary scale.
Electricity systems under pressure. Extreme heat. Congested cities. Industrial inefficiency. Agricultural productivity. Water. Logistics. Healthcare access. Affordable cooling.
These are usually described as constraints on development.
They are.
But they are also sources of information.
Someone living inside an unreliable system knows where it fails.
An engineer who has spent years inside a factory sees inefficiencies that an outsider does not.
A founder working alongside farmers understands where an elegant technology becomes economically useless.
A power-system operator may understand a bottleneck long before it appears in an investment thesis.
This kind of knowledge is rarely treated as an asset.
Perhaps it should be.
Proximity to a problem can itself be a form of intelligence.
The innovation economy is extremely good at attaching enormous valuations to knowledge once it has been converted into intellectual property, software or a company.
It is considerably less sophisticated at recognising the tacit knowledge that precedes them.
That is where India's opportunity becomes interesting.
Technology's centre of gravity is also shifting.
AI may appear weightless, but its consequences are anything but.
Computing requires data centres. Data centres require electricity, cooling, land and transmission. Electrification requires grids and storage. Advanced manufacturing requires materials, machinery and energy.
India's electricity demand is forecast to rise by an average of 6.4% a year between 2026 and 2030. The International Energy Agency expects the country to add more than 570 terawatt-hours of annual consumption during that period. Cooling alone could account for more than a fifth of demand growth.
These numbers are usually presented as an infrastructure challenge.
An entrepreneur might read them differently.
Five hundred and seventy terawatt-hours of additional demand means an enormous market for generation, storage, grid intelligence, efficiency, cooling technology, industrial optimisation and demand management.
A constraint of sufficient scale eventually becomes a market.
And a constraint shared by enough countries can become an export market.
India is hardly alone in confronting rapidly growing cities, rising cooling demand, constrained infrastructure and the simultaneous need to industrialise and decarbonise.
A technology engineered to work affordably under Indian conditions may therefore have relevance far beyond India.
This is a much more ambitious proposition than becoming a cheaper version of Silicon Valley.
India must stop importing the definition of innovation.
There is, however, an uncomfortable side to the argument.
Having difficult problems does not automatically generate important companies.
Nor does engineering talent.
An ecosystem must be capable of recognising promising solutions before their value becomes obvious.
That is harder than it sounds.
Capital tends to follow familiarity. A business model that has already succeeded in America feels easier to finance. A founder from a recognised university looks easier to assess. A fashionable AI category can attract attention faster than an obscure industrial technology that may take years to prove.
The result is an irony.
India could possess unusually valuable information about the problems of the future while continuing to allocate capital according to the investment patterns of the past.
Government policy is beginning to acknowledge part of this mismatch. India's ₹1 lakh crore Research, Development and Innovation scheme is intended to support high-risk, high-impact technologies, while the newer ₹10,000 crore Startup India Fund of Funds 2.0 specifically includes deep technology and innovative manufacturing among its priorities.
The funding architecture matters because many of the technologies India most needs will not behave like conventional software businesses.
A grid technology needs pilots.
A new industrial process needs demonstration facilities.
Advanced materials may need years of development.
Energy technologies may eventually require infrastructure finance rather than another venture-capital round.
India, therefore, needs more than additional startup capital.
It needs investors, corporations and institutions capable of recognising different forms of technological value.
There is a question investors routinely ask founders: Can this scale?
It is a necessary question.
But perhaps another should precede it: What does this founder understand that the rest of us have not recognised yet?
The answer may be a scientific breakthrough.
But it could equally be a mundane observation about why an industrial process wastes energy, why a supply chain repeatedly breaks or why a technology designed for wealthy markets fails when deployed somewhere else.
The world's most valuable companies are not always created by inventing needs.
Often, they are created by seeing an existing need differently.
AI makes that ability more valuable, not less.
If machines become increasingly capable of generating code, analysing information and producing answers, one of the enduring human advantages may lie further upstream: Choosing which questions deserve attention.
AI can process what has been recorded.
Much of the world's useful knowledge has never been recorded.
It lives in experience.
I returned from San Francisco admiring something Silicon Valley still does extraordinarily well.
Not technology.
Confidence.
It has created an ecosystem in which a small group of people can encounter an unusual idea and imagine, remarkably quickly, that it might become a very large company.
India should borrow that confidence.
But it need not borrow Silicon Valley's problems.
It already has better ones.
The country's next technology advantage may, therefore, depend less on whether it can produce another generation of familiar unicorns and more on whether it can build institutions capable of noticing unfamiliar value: Entrepreneurs outside conventional networks, knowledge emerging from difficult environments and technologies whose importance is not yet reflected in a fashionable category.
India's problems are not, by themselves, a competitive advantage.
The ability to recognise the intelligence hidden inside them could be.
And in an age when technology is becoming easier to build, that distinction may matter more than we think.
(The views expressed are personal)
This article is authored by Gayatri Panda, investor and founder, Indian Tech Society, London.