Silicon Valley’s builders are building faster than ever
Silicon Valley is miles ahead of the rest of the world once again – and what this means for everyone else.
I’ve been going to the Bay, and Silicon Valley in particular, for over 25 years. As an insider-outsider I’ve seen it change through each subsequent technology transition. During my April 2023 visit, I felt the Valley change once again as AI was becoming the new computing platform. I wrote on this occasion:
Silicon Valley is never more vibrant than when there is a new platform.
Since then, I’ve been back three time, most recently last week. On this latest trip, I met more than 30 founders and eight VC firms covering the gamut from semiconductors through to hard-core AI research and more mainstream AI applications. I got more than enough anecdotes to support the acceleration argument. But we also see it in the data. Companies, like Cursor, have grown their revenues 100x in just a year. It is unprecedented.
It’s become clear to me that Silicon Valley is miles ahead of the rest of the world once again. Ideas travel faster today than ever before, but it still takes time for the vibes to cross continents and oceans. That’s why it is critical to pay attention to what is being invented in Northern California and how behaviours are changing in tandem with those innovations.
Why is this happening?
To misquote political strategist, James Carville, “it’s the technology, stupid.”
From the early days of semiconductors and mainframes in the 1960s and 1970s, through the personal computing boom of the 1980s and into the internet and mobile revolutions of the late 1990s and 2000s, the Bay Area has repeatedly raced ahead of the rest of the world. Each time, new technologies—microprocessors, graphical user interfaces, online commerce, social media—were conceived and refined here before spreading globally. These waves established Silicon Valley as not only an engine of innovation and thus a magnet for talent and capital.
Most recently, following the success of Facebook and the iPhone in the early 2000s, the gap between the US West Coast and the rest of the world narrowed. Partly because it’s easier to make progress from a standing start. The UK got its Silicon Roundabout (where my last startup was based) and the ecosystem of founders, incubators (like Seedcamp and Entrepreneur First) and maturing venture firms. Zero to something is a big step.
Another reason for the narrowing gap was that a lot of the value on mobile and social was higher up the stack, at the consumer surface, eyeballs and advertising, two areas where New York and Europe have relative strength. Of the six largest advertising groups in the world, three are headquartered in Europe and two in New York.
ChatGPT has energised the patchwork of communities that exist or are returning to the Valley. The Bay Area feels further ahead of the rest of the world in a way it hasn’t since the late 1990s. AI is a nerd’s dream in a way that mobile and social never were. It demands new semiconductors to handle massive computation. It requires novel memory architectures to process vast language models. It needs breakthroughs in algorithms and deep computer science - the kind of technical stack-building that Silicon Valley was originally built for. The opportunity of AI is about fundamental computer science and that’s when Silicon Valley pulls ahead.
How do we know it’s real?
Because of people, growth and money.
Pre-ChatGPT gravitational pull, heightened by the pandemic, moved attention away from the Valley. Between 2018 and 2022, several high-profile figures relocated to Miami, Austin and other cities. Miami attracted big names like Keith Rabois, Shervin Pishevar and David Blumberg. Austin seduced Elon Musk, Dropbox’s Drew Houston and Palantir’s Joe Lonsdale. At the same time, large technology companies (not merely big tech firms, but others like Oracle) started moving operations from the Bay Area and engaging in layoffs. By the numbers, these layoffs are big. A single round of layoffs at Meta in November 2022 would represent 3,800 Y Combinator founding teams (assuming a founding team size of 3).
But large companies aren’t what makes the innovation, what reshape industries. It’s (mostly) small companies, led by visionary, quixotic instigators that do.1
Location matters yet again. Even in a digital world with tools that support distributed teams, from GitHub to Slack to Zoom, being in the same physical space seems critical. Agglomeration, clumping together, matters. Engineers working together, sharing insights face-to-face, in the same room. In large organisations where change is slow and process defines everything, remote working can work. For innovation, despite obvious successes (like MySql and, perhaps, HuggingFace), colocation counts.
Silicon Valley seems to be in the lead with AI talent. I conducted a highly imperfect exercise to estimate how deep AI talent runs in a number of major metros. I searched LinkedIn for people with “Artificial intelligence” in their bios. It's imperfect because anyone can put “AI” in their bios and many people with exceptional AI credentials do not (one example is Amjad Masad, founder of Replit, the AI coding platform).
The Bay Area has more people sharing more ideas and how-to. This creates competitive environments to be more creative and productive. And this isn’t a new phenomenon. Over a more than three-decade period to 2008, the Bay Area’s share of patent filings in the US rose from 4% to 16%, while it’s relative population remained unchanged.
The office real estate market in Silicon Valley had its best quarter in years, with rents rising slightly and the vacancy rate dropping in the fourth quarter, after years of increasing emptying.
The business
We are reasonably familiar with the data from major AI companies below. That two grew revenue 100-fold to exceed $100 million in a year is quite astonishing.
From what I heard in discussions with various investors, it isn’t just the “household names” enjoying such exponential leaps in revenues. The pattern holds in earlier-stage companies as well.
One pointed out that in a pre-genAI world it used to take an exceptional team of six or seven about a year to get to a million dollars in annual recurring revenues (ARR). Today, he is seeing teams of two founders get to the milestone in under three months. Founder productivity has been multiplied by using AI tools for coding, design, testing, marketing and sales, and the huge demand for these products has concertina’d timelines.
This isn’t surprising. Software engineering emerges as the dominant sector for AI integration, accounting for 37.2% of queries on Claude. And startups more than any type of company, derive their customer revenue from software.
(To be sure, there are many firms outside the Bay Area growing increasingly rapidly. Lovable, one of my new favourites and based in Sweden, has surpassed $4 million in ARR within four weeks of its launch in November 2024.)
But the Bay Area is also a place with many early-adopting personalities, those types of people to jump into an Uber (or Waymo) ahead of those in other cities, making for an appealing market for early sales.
The Valley’s still got its groove, a cycle that feeds itself.
The money scaffold
The final dimension is the increasing availability of the right kind of capital to support growing ventures.
North America makes up over half of total global VC funding; the Bay Area is a quarter of that. The proportion is higher than in the past decade.
And globally, AI is attracting an increasing share of VC funding. In the US in 2024, 42% of all venture capital went to AI companies. In the Bay Area that was 71%.
In addition, we’re in a winners-beget-winners cycle where companies are raising mega-private rounds—most of the capital winding up in the Bay Area, including in Databricks, OpenAI, Anthropic, xAI, Waymo, Scale and Perplexity.
The self-reinforcing cycle
Prior to ChatGPT, Silicon Valley was casting around for its new platform: would it be metaverse or crypto or some horrible combination of both.
But in just over two years, the entire ecosystem has reoriented around AI. And this is what Silicon Valley does best: an ecosystem that is both resilient and responsive to technology signals. Being big and fast normally involves tradeoffs – think a powerlifter vs. a sprinter. Not so here.
The situation is not as monochrome as it was back in the late 1990s. For one thing, other ecosystems are more mature than they were previously. At the same time, the range of areas we’re exploring from materials to the microbiome, from batteries to biologics, from drones to humanoid robots and new materials, is broader. Some of those deeper tech areas have strengths outside the Valley.
And it might be the necessity — especially around defence — that will breed deep innovation. Consider one of Europe’s hottest (AI) startups, Helsing, which gained momentum amidst the war on Ukraine and the declining Atlantic alliance between the US and Europe.
Indeed, I have invested in geographies outside of the Bay Area. Of my most recent 20 investments, I’ve got founders in Cincinnati, Washington, Houston, Chicago, Paris, London and Berlin and only five deals in the Bay Area. But three of my last five investments are in San Francisco—and most of the deals I’m now considering are also there2.
Broader implications
Consider what this means… The Valley’s acceleration in AI may mean that more of the core technical decisions about AI architecture, capabilities and safety mechanisms are concentrated in one region. Admittedly, there is a wrinkle here as Google’s Demis Hassabis is based in London, but Google is largely a Bay Area firm. And the self-reinforcing cycle in Silicon Valley means other regions may fall further behind in deep tech capabilities. This isn’t just about consumer applications; it’s about fundamental AI infrastructure, research and compute.
What does it mean for other regions, excluding China?3
India
The most interesting of these is India. It presents a fascinating counterpoint to Silicon Valley’s AI acceleration. With GDP growing at 7%+ annually and a digital-first economy of 1.4 billion people, India might not need to compete directly on Silicon Valley’s terms. Instead, it’s building its own distinctive technical stack born from solving distinctly Indian challenges.
Less than a decade old, the Universal Payments Infrastructure, now handles roughly 15 billion transactions monthly, more than half of India’s digital transactions. By transaction count, UPI is now bigger than Mastercard worldwide. This is an achievement unimaginable in the US.
This points to how India might approach AI: not by competing to build foundational models, but by solving unique problems at unprecedented scale — whether it’s multilingual AI for India’s 22 official languages or AI-powered financial inclusion for the world’s largest middle class.
The next wave of Indian innovation will likely come from combining AI with its digital public infrastructure. Just imagine AI layered on top of UPI, Account Aggregator frameworks and the health stack. They can do this.
Europe
The picture is different for Europe. On the one hand, Europe is waking up, focus on innovation and reindustrialisation, and getting rid of things that block the entrepreneurial journey, be they regulatory or other. As Mario Draghi, the former Prime Minister of Italy, said in a powerful speech to European politicians, “you can’t say no to everything.”
European AI firms, like Mistral and Kyutai, should emphasise open-source innovation. Open-source innovation levels the playing field and puts competitive pressure on closed-source firms. That pressure will be most keenly felt on their margin structure. Open-source, whether from Mistral or DeepSeek, frankly, allows anyone to build on collective strengths.
The continent is about to get a huge injection of cash for defense. This is an opportunity to take a Day One mentality and use defense dollars euros to catalyse breakthrough technologies. European defense budgets have increased by roughly €120bn annually since 2020 and look likely to rise further. Helsing seems like in might be one early winner. There should be others.
Europe could be well positioned in industrial and scientific applications, with the deep heritage of firms like Siemens and ABB. Although it isn’t the heritage that really matters, it is the data they have about industrial systems. There are tons of other niches — which can turn into huge markets — where Europe can producer winners: ARM, in semiconductors, being an obvious example.
The coherent thread is that Europe can compete best where openness, deep expertise, and infrastructure matter more than pure scale or consumer reach. Open-source serves as a multiplier for Europe’s other advantages. It allows industrial expertise to be shared, infrastructure to be standardised and talent to be pooled across borders.
Meet me in the Valley
Silicon Valley’s resurgence in the AI era is fundamentally about the rebirth of what made the Valley special in the first place. While the mobile and social eras distributed innovation globally, AI has brought us back to fundamental computer science, the Valley’s home turf. If you truly want to understand what the next technology frontier is about, you’ve got to spend time there.
At the same time, there’s a distinctive sense of broader opportunity space that will take shape over the coming months and I’m excited to see it happen. Other regions shouldn’t replicate Silicon Valley’s playbook exactly as Europe and other ecosystems tried to do in 2010s. Some things like more sensible labor laws, lower administrative burdens on company formation and reporting, better tax treatment and greater access to capital do need to be imported with alacrity.
But a different path may work for these other places. India is building on its digital public infrastructure, while Europe can leverage its industrial expertise and open-source innovation.
Of course, while all this happens, Silicon Valley’s builders will keep building.
As I wrote this, Microsoft announced a quantum computing breakthrough.
Top-tier European AI founders are always welcome to get in touch!
China is a different ball game. Will turn to that in due course.
Great seeing you in SF and love your observations. My only other observation is that it strikes me that AI is crossing through all sectors and industries - in a way that is more profound than the advent of the internet. I'd also add that even as someone who has lived in Silicon Valley a long-time, the pace of change is staggering.
Great article! I started in the VC Debt business in the valley in 1986 and lived through a number of cycles over the decades. Your insights are spot on.