(The author, Kabir Narang, is a global tech investor and a founding general partner at B Capital. Views expressed are personal.)
As I look back on the last 25 years in tech and investing, I am struck by how much has changed and how much hasn’t.
In 1999-2000, I was a graduate student writing for The Oxford Student newspaper while studying for my Master’s. It was the height of the dot-com boom. Silicon Valley was euphoric, valuations were sky-high and it seemed every other week another startup was rushing to an IPO.
Fast forward to today: we are in the midst of an AI and robotics revolution that will dwarf the internet and mobile revolutions in its impact.
Today, transformative themes—generative AI, autonomous systems, robotics, and dual-use defence tech—aren’t just buzzwords; they are fundamentally re-architecting industries, business models and global economic priorities in ways that far outstrip the pacing and reach of the dot-com era.
What’s different is that, unlike 25 years ago, these innovations are already delivering tangible enterprise value in some cases (but that’s a plug for another article) and real-world adoption, with the backing and executional rigour of both tech giants and startups—a convergence we simply didn’t see during the original internet boom
Trillions of dollars are being invested. Yet, certain fundamentals and patterns from 25 years ago persist. Here are a few observations on what has changed and what hasn’t in the world of tech investing.
The dot-com bubble: Then and now
At the turn of the millennium, the dot-com mania was in full swing. Nearly a thousand companies went public in the span of just two years—565 IPOs in 1999 and another 400+ in 2000. Most of these were internet startups riding the hype. Moreover, capital was abundant—venture investments exploded and anyone with a bold dot-com idea could attract funding.
By March 2000, the Nasdaq index peaked, and the celebration climaxed with the now infamous ‘dot-com Super Bowl’ in January 2000, where 14 internet startups bought Super Bowl ads.
The party ended abruptly: within two years, the Nasdaq plunged around 78% from its peak, wiping out countless companies.
Yet, amid the wreckage, a handful of winners not only survived but thrived. Companies like Amazon emerged from the dot-com crash stronger and went on to dominate. Another survivor, eBay, persisted while hundreds of lesser-known e-tailers disappeared. The famous venture capital ‘power law’ – where a few outliers account for the vast majority of returns was vividly demonstrated.
I have witnessed the power law assert itself repeatedly, across geographies and stages. Spotting and backing these compounding winners has also become my defining pursuit—because in venture, success is anything but evenly distributed.
The big lesson: technological revolutions tend to overestimate short-term winners but underestimate long-term giants.
Technological revolutions tend to overestimate short-term winners and underestimate long-term giants
AI–bigger atage, bigger stakes
Today’s environment with AI and deep tech creates a sense of déjà vu, and in other ways is fundamentally different. The stakes and scale are even larger now. By many accounts, AI represents a transformation 10 times bigger than the Industrial Revolution.
As an investor, I feel the familiar mix of excitement and anxiety. On the one hand, there’s exuberance: everyone is chasing the next AI unicorn, and venture capital money is pouring in.
On the other hand, memories of 2000’s crash instil caution. We know how hype cycles can end. The good news is that today’s leading tech companies are far more substantial than most 1999-era dot-coms. Many AI-driven firms already have real revenues, and the big players incorporating AI—the likes of Google, Microsoft, Amazon, and Apple—are highly profitable.
Companies chasing AI today generally have stronger fundamentals. The excesses are a bit more restrained—speculative frenzy is present, but “far less extreme” than during the bubble in terms of valuations and business quality.
Companies chasing AI today generally have stronger fundamentals.
Yet some things never change. Investor psychology continues to oscillate between the euphoria of fear of missing out and the anxiety of potential losses. In 2021, we witnessed a historic IPO wave, with more than 2,700 companies going public globally and over $600 billion raised—figures that dwarfed even the 1999-2000 frenzy.
That exuberance quickly gave way to reality as a few marquee tech IPOs faltered, and the market for new listings sharply contracted in 2022 and 2023. This is a vivid reminder that the boom-and-bust rhythm is an enduring feature of public markets.
The context may differ, but market cycles rhyme with history—or as Mark Twain said, “History doesn’t repeat itself, but it often rhymes.”
This year, Mainland China and Hong Kong have reasserted their dominance in the global IPO scene. Greater China now accounts for about one-third of total global IPO proceeds, and Hong Kong has reclaimed its position as the world’s top hub for new listings, fueled by policy reforms, liquidity and a surge in mega offerings from sectors like tech, EVs, and healthcare. The recent momentum in China is a reminder to investors how swiftly fortunes can shift on the global stage.
What’s Changed? Harder tech, smarter money
One striking change in the last quarter-century is how technology itself has deepened.
The late ‘90s were about connecting people via the internet. Today’s frontiers—artificial intelligence, robotics, biotech, space—are more capital-intensive and technically complex.
It’s serious business models now: enterprise AI startups selling to Fortune 500 companies, robotics firms with real manufacturing, even defence-tech companies working on hypersonic missiles or autonomous drones.
These aren’t just two founders in a garage building a website. Often, they require large R&D budgets and patient capital. The encouraging sign is that investors, having been burned before, are performing more due diligence. Quality of revenue matters, and investors ask tougher questions (though exuberance can still override caution at times).
Another change: the geography of innovation has evolved, yet also concentrated. In 1999, Silicon Valley was the undisputed centre of the tech universe, with a nascent scene in places like Seattle, Boston, and some activity in Europe.
Today, innovation is more geographically spread—from Bangalore to Tel Aviv to Toronto. However, the gravitational pull of a few hubs remains incredibly strong. In fact, the AI boom has been highly concentrated in just a few regions. Recent data show that almost 80% of venture funding for core AI startups went to just three places: Silicon Valley, Beijing, and, surprisingly, Paris.
80% of venture funding for core AI startups went to just three places: Silicon Valley, Beijing, and, surprisingly, Paris.
Silicon Valley alone took about 65% of all AI startup funding over the last 18 months—an astounding share. This mirrors what I call the ‘physics forces’ of tech: talent, capital, and expertise tend to attract each other, creating dominant hubs.
Two and a half decades later, despite all the talk of a flat world, Silicon Valley and China’s top tech cities are even more central to cutting-edge innovation. Every aspiring AI founder I meet from an ’emerging’ ecosystem ultimately ends up flying to Silicon Valley for inspiration or access specialised talent. The network effects in tech are real and seemingly growing.
What hasn’t changed: Human nature and power laws
While technologies evolve, human nature in markets remains constant. Greed and fear, hype and panic—these undercurrents drove the South Sea Bubble in the 1700s, the dot-com bubble in 2000, and arguably parts of the crypto craze recently.
As an early writer during my Oxford days, I interviewed business leaders, movie directors, and even famous cricketers visiting Oxford Union debates, and I noticed how storytelling and optimism often trumped sober analysis when people fell in love with a trend. That hasn’t changed. We still get carried away with narratives—be it “the internet will change everything” in 1999 or “AI will transform everything” in 2025. Both statements are true to an extent, but exuberance can lead to indiscriminate bets.
Crucially for investors, the power law of outcomes remains a defining feature of tech investing. In venture capital, a handful of companies will continue to generate the bulk of returns. We might place hundreds of bets on AI startups but it’s a safe bet that five to ten years from now, only a few will dominate the market or IPO at massive valuations.
We might place hundreds of bets on AI startups but.. only a few will IPO at massive valuations.
The rest will either fail, get acquired or slog along. The big winners today might be names we can already guess (the mega-cap tech firms are pouring billions into AI) and / or they might be new entrants that follow the trajectory of an Amazon, Meta or Google from the previous generation.
As an investor, this power law means you must be prepared for a high failure rate in your portfolio, but those handful of big successes will compound and more than pay for all the duds. It also means one must think big—incremental ideas rarely produce the outsized outcomes that justify the risks.
