For 30-odd years, Silicon Valley had an extraordinary run. American entrepreneurs pioneered innovative digital businesses that disrupted sector after sector of the economy. These startups were massively profitable, minting enormous new fortunes and a new generation of titans.

But that era is over. New forces are eroding the traditional role that startups played. The demise of globalization, the return of geopolitics, the completion of digitization, and the rise of AI have tilted the balance back in the favor of established players. AI is now dominated by vast companies like Microsoft, Google, Meta, and Nvidia. That doesn’t mean startups are dead, however. Rather, they and their investors have to change the way they conceive of their role. The old goal was to disrupt. Now it must be to transform.

In the first era of internet-driven innovation — the “dot com” boom — companies made their mark by bringing services (mostly shopping) online. Although that first boom went bust, the digital infrastructure it created kept expanding. As it did, entrepreneurs were able to start digital-first businesses: new banks, insurance companies, travel companies, and healthcare providers. Other innovators, meanwhile, started companies intended to allow the legacy businesses to stay competitive in this new digital reality — for instance, the cloud-computing startups Snowflake and Datadog.

When startups actually disrupted an industry, it was because they unlocked both a software advantage and a business model one. Category-defining companies like Airbnb (a General Catalyst portfolio company), Uber, and DoorDash thrived by offering a new digital interface that also revolutionized the way people used a real-world service. But other industries remained fundamentally unaffected: the past 30 years of innovation didn’t, for example, create a real rival to J.P. Morgan or State Farm. Indeed, in many sectors, the incumbents only grew stronger in the digital era, immune to the disruption emanating from Silicon Valley. Still, there were enough disruptors for venture capital to enjoy extraordinary returns.

It wasn’t just American ingenuity that fueled this era; broader global trends also helped. The mid-2000s saw the birth of the truly digital consumer, with the “rise of the rest” having created rising middle classes across the planet. The global rise of Facebook and smartphones attest to this phenomenon. But globalization was reaching a peak, with global tariffs finishing their long decline to record low rates. Its chief beneficiary, China, took advantage of its WTO membership and became the world’s factory. Living through this age, entrepreneurs adopted a mental model that was post-historical and post-geographical. Thanks to globe-spanning supply chains, widespread online access, and affordable cloud computing, they could serve consumers digitally in every industry in every country.

Venture capitalists, for their part, settled on a winning strategy: as long as a company had a product that the market wanted, the goal was to quickly gain market share and become the dominant player in a given category. Hence the “move fast and break things” mantra that generated so much money and societal turbulence over the last two decades — not just in new industries like social media but also in traditional ones like healthcare.

When the Great Recession hit, startups were helped more than they were hurt. Although they suffered like other sectors from the immediate crash, they went on to benefit handsomely as the Fed increased the money supply, fueling a new era of cheap capital. (From 2010 to 2014, the amount of funding that the venture capital industry raised more than doubled.) In a similar way, the Covid-19 pandemic also supercharged startups. It wasn’t just that even more of everyday life moved online; it was also that investors took advantage of governments’ stimulus packages to rush through new deals, with fresh capital injections sending stocks to new highs and expanding the amount of available capital.

But the most important effect of the pandemic was to bring to the surface the anti-globalization fervor that had been rising for years. By 2020, tariffs were already back in vogue, thanks to the Trump administration, and by the World Bank’s count, countries were signing 60% fewer trade agreements than they had been two decades earlier. The pandemic laid bare the risk inherent in globalization: countries that lacked the ability to manufacture their own vaccines or masks had to wait behind others for seconds. Then came Russia’s 2022 invasion of Ukraine, which heightened the focus on security and self-sufficiency. The rise of U.S.-China tensions has raised the prospect of a world split into two economic zones.

These geopolitical headwinds have hurt startups more than they have hurt established players. For one thing, as governments seek to secure supply chains, they are turning to stalwarts, not startups. Think of the firms that stand to benefit most from the CHIPS and Science Act, which provides $52.7 billion in subsidies for semiconductor R&D and production. It’s not the unheard-of new entrants that are getting the money. It’s the stable, long-time companies that can offer surety of supply: Taiwan Semiconductor Manufacturing Company, Samsung, Micron Technology, and Intel.

For another thing, the fracturing of geopolitics has made government policy newly relevant. In the previous era, companies could grow in a relatively regulation-free environment. (Uber, for example, largely decided to ignore regulations governing the taxi industry, hoping — correctly — that its service would be so popular that the law would eventually catch up.) But the past few years have seen a mushrooming of new trade restrictions, domestic subsidies, and privacy laws that companies must comply with or seek to change. And the companies most skilled at doing that are the big, established ones. Last year, Amazon and Meta each spent nearly $20 million on its American lobbying operation, far beyond the means of any startup.

Making matters even worse for digital startups, they have to contend with market saturation. In the West, most people are now digital consumers. By now, most businesses have digitized nearly all their processes — especially since the pandemic began. More than 90% of companies, for example, have adopted cloud computing.

How AI Fundamentally Alters Startups’ Nature

But just as this technological trend has run its course, a new, far more disruptive one has arrived: the rise of artificial intelligence. For the first time since the advent of the internet, every CEO in every industry in every country is trying to figure out how to adopt a technology at the same time. This time, ambitions and hopes are even higher. Not only does AI offer to translate existing processes from one realm to another, as digitization did before it, but it also learns by itself.

That key difference means that the productivity gains we can expect are not a one-time “step change” but rather a continuously improving “slope change.” At its core, AI is a workforce-transforming technology, one that will unleash human productivity by creating a parallel labor pool that can shoulder the burden of much of the work humans would rather not do. AI will become a source of abundance, exerting deflationary pressure across the economy by increasing the supply of labor for caretaking, tutoring, maintenance, and more.

AI represents a far bigger opportunity than anything the tech sector has ever faced. But it requires Silicon Valley startups to shift their mindset, to no longer seek to disrupt and demolish established firms but to transform them. That’s because startups are in many ways at a disadvantage with this new technology. Successfully harnessing AI requires two things: lots of data and lots of expensive computing power. Big companies have access to both. They have the data on which to build their models, the money to pay for all the computing power to analyze it, and the customer relationships to immediately monetize these expensive endeavors.

Take code generation, the core task of all software engineering. The first company to dominate this space with AI was not some scrappy startup. It was Microsoft, the 49-year-old company with a $3 trillion market capitalization. Its GitHub Copilot, the most popular AI-powered code completion tool, now boasts over 1.8 million paid subscribers.

There’s no reason to believe yet that the AI race will be winner-take-all; there is still room for scrappy startups to eat away at market share, and indeed, many promising ones are tackling code generation. But it will be harder to outpace the big companies in terms of the quality of their tech and the quantity of their data and resources.

That said, with demand for innovation higher than ever, startups will still have much value to add. Where their new opportunity lies is in finally performing the services themselves, rather than merely enabling the work other companies do. For years, Silicon Valley has been obsessed with introducing the efficiency gains of software to every commercial activity — creating a digital interface for ordering from restaurants, for example. But in the age of AI, software isn’t merely an intermediary for a service; it is the service.

Consider the customer service call center. In the old era, startups merely sold the software used by employees at large companies’ call centers. With AI, the call center startup provides the service itself — namely, the chatbot that customers are talking to. This is unfamiliar territory for software startups, which are accustomed to operating with relatively few assets and enjoying profit margins as high as 80%, as opposed to the much lower margins that prevail in the service sector.

But AI will help close the gap in profit margins, and the potential market size for startups that can make this transition is in fact several times larger, because they can go after the entire value chain while remaining at their core a technology company. Indeed, despite all the hype around the tech sector and its prominence in American discourse, it is responsible for only about a tenth of U.S. GDP. If tech entrepreneurs seize the new opportunities, performing services and not just shaping workflows, then that share could get much larger.

How can American entrepreneurs tap into the new startup advantage? Startups will have to collaborate much more closely with existing enterprises if they want access to their data, expertise, and customers. They will also have to develop a new understanding of what it means to provide services, even if they are digital or performed by AI agents. That means making new pricing models, providing customer support, and designing products for end users, not just the enterprises acting as intermediaries.

Perhaps most important, they will have to embed the concept of responsible innovation at the core of what they do, thinking constantly about the effect that their technology will have on workers in the sectors they are transforming. If AI takes over 25% percent of a person’s job, what happens to that newly spare capacity? In healthcare, you can imagine nurses using this extra time to shift from reactive care to proactive care, potentially reducing future patient illnesses. But making that shift requires strong leadership willing to invest in strategic decisions. There will always be a short-term temptation to reduce workforces, but succumbing to it will lead to a long-term loss: the companies that will ultimately win out will be those that reinvest their talent to higher-value activities.

Startups still have many advantages that big companies can never dream of obtaining. They attract risk-seeking talent. They act quickly. They innovate. They adapt. For all the new headwinds, those characteristics should serve them well as they navigate the transition from the last era of innovation to the next one.

https://hbr.org/2024/10/ai-is-transforming-the-startup-landscape