How many startups are at UC Berkeley? The University claims over 1,600 — a number it plasters endlessly in brochures, articles, and alumni reports.
I remember reading and re-reading those brochures, desperate to fall in love with the school I did not expect to go to. Those who come across that figure might expect a community of hardcore founders betting their entire livelihoods on realising wonderful, fantastical dreams. It’s a magnetic expectation. Every year, it rips an untold number of impressionable idiots away from comfortable careers in courtrooms, hospitals, and offices, and down the yellow brick path through Sather Gate. I’d know because I’m one of them.
Yet if there really are 1,600 startups at UC Berkeley, where are they?
The Britannica Encyclopedia characterises a startup as a fledgling business, especially one that intends to grow very large, very rapidly. Paul Graham, the founder of Y-Combinator and arguably the most influential venture capitalist today, famously wrote “startup = growth.”
Graham’s “growth” definition is more demanding than it first appears. Bakeries, law firms, consulting firms, or local service businesses can be profitable and grow, but they are not startups. This is because they grow through known models in known markets. They may expand, but they do so predictably and incrementally.
A startup, by contrast, must aspire to expand unpredictably and rapidly. To do so, they cannot compete against mature firms in mature markets using mature strategies. In such markets, incumbents already control customers, distribution, and scale, forcing new entrants to grow slowly by taking market share bit by bit. Growth achieved this way is incremental, not exponential. For a startup to grow quickly, it must be different from the incumbents in some fundamental manner, creating new markets or redefining existing ones.
To redefine a market, a startup must first believe the market (or more broadly, the world) is wrong about something. It must act on a claim about how the world should work that others do not yet accept. In doing so, it creates a kind of nascent ideology, and I don’t mean in the strictly political sense. I mean a founder’s structured belief about how the world works, what it lacks, and why others will eventually come to see it differently.
Apple began with an ideological claim. It was Steve Wozniak, Steve Jobs, and Ronald Wayne’s unsubstantiated thesis that the computer would become the desktop appliance, as an extension of the human mind. So did Facebook. It too, was birthed of its founders’ untested conjecture that people would want their real identities and relationships to exist on a single online network. Every quintessential startup did not begin primarily as spreadsheet models or market analyses. They began with arguments.
In this sense, an early startup is like an intellectual thesis. It is a claim about reality that must survive confrontation with the world.
If Berkeley truly produced more than 1,600 startups each year, the implications would be extraordinary. A university capable of generating that many attempts to redefine markets would dominate the technology industry. Its founders would routinely produce the next generation of category-defining companies.
Yet the evidence suggests the opposite. When measured by the percentage of those “startups” that become billion-dollar, category-defining companies, Berkeley lags far behind schools like Stanford, producing fewer than half as many despite enrolling more than four times as many students.
There’s a myriad of excuses for this discrepancy. Many have been convinced that Berkeley students are simply less talented than their counterparts at Harvard and Stanford. The reality is that Berkeley’s engineering and computer science programs are among the strongest in the world. Our undergraduate programs in computer science, business, and engineering frequently outrank those at highly selective private institutions.
Others point to a lack of opportunity. Yet, Berkeley sits less than an hour from the same venture capital firms and technology companies that surround Stanford, and has close, accessible public transit to downtown San Francisco, one of the most venture capital and technology-dense regions in the world.
Berkeley’s claim to produce so many startups each year raises an odd question. Not why Berkeley produces so many startups — but whether most of them are startups at all. Many students call their small business ideas “startups” to get into prestigious clubs. Researchers are incentivized to call their projects “startups” to secure additional funding. The “startup” is an all-consuming word. It means everything at UC Berkeley, and therefore, nothing at all. And the reason lies in Berkeley’s startup ecosystem. There are three structures I’ve noticed in particular.
The first is tied to social selection. Dave Chapman wrote a wonderful essay called “Geeks, MOPs, and sociopaths in subculture evolution.” In it, he described how subcultures change as they grow. Small communities begin with people genuinely obsessed with the activity itself — the “geeks.” But once the activity becomes visible and prestigious, it begins attracting others who are drawn less by the work than by “mopping” up the status attached to it. Chapman calls these newcomers “MOPs.”
Something similar has happened to startups at Berkeley. Starting a company here is no longer simply an act of entrepreneurial obsession. It has become a form of social currency. Entrepreneurship clubs are among the most competitive organizations on campus. Students list “founder” on LinkedIn long before they have customers. Venture capitalists visit campus, host dinners, and fund student projects because everyone knows: building a startup is impressive.
Once that signal exists, the ecosystem begins attracting people who want the recognition associated with startups rather than the work itself. These founders tend to follow one of two paths. Some choose problems based on opportunity — popular markets, fashionable technologies, ideas that sound impressive in a pitch deck — all things that inherently force you into following a recognized trend.
Others stall much earlier. Many would-be founders fall victim to the honeypot of prestige. Clubs, incubators, and programs are often exclusive — and fun. There are free concerts and trips to Hawaii. The act of joining the club, forming a team, and pitching the idea already delivers the recognition they were seeking.
The “MOP” invasion at Berkeley greatly dilutes the culture, but it does not directly kill existing startups. That is left to our second structure: institutional translation. The confluence of incubators, clubs, academic programs, and accelerators extracts any nascent ideology and replaces it with uninspired shareholder capitalism, in a gradual but well-intended process.
Most worthwhile ideas begin as something strange and difficult to articulate, something akin to a dream. It is often personal, sometimes irrational, and rarely optimized for presentation. But the problem is that institutions inherently require new ideas to be reformulated and squeezed into a narrow range of legibility to be categorized and evaluated.
Clubs teach students how to package ideas for social recognition.
Accelerators teach them how to make ideas legible to investors.
Incubators teach them how to fit ideas into existing venture categories.
At each stage, the idea becomes more polished and less ideological. It becomes less of an extension of the founder’s humanity and more of an arm of a machine in pursuit of popular industry trends. The founder learns which arguments are rewarded and which are ignored: to follow trends rather than create them. The original belief that motivated the project is gradually replaced by something more legible: market size, competitive analysis, and financial projections.
This transformation produces ventures that are easier to fund, easier to evaluate, and easier to understand. But it also removes the very element that made them startups.
The third structure kills ideas before they begin. I call it ideological stagnation, and I first understood it through Pete Flint. Flint has founded two billion-dollar “unicorn” companies, the most recent of which went public on the New York Stock Exchange before merging with Zillow in a $3.5 billion deal. He now runs a $1.5 billion venture capital firm and is in the unusual position of observing today’s founders while remembering what it took to build his own companies.
During our conversation, Flint asked me to consider two hypothetical startups: one attempting to solve a grand, widely recognized problem such as world hunger, and another focused on something obscure like data transfer latency.
“I see one founder pitching his idea to change the world and solve world hunger,” he told me. “He presents, and everyone says, ‘That’s awesome. Well done. Love it.’ And then he feels validated.”
Flint contrasted that with a very different response to a more technical idea.
“And then you have another founder pitching a startup that reduces data latency in data-based queries,” he said. “People react with, ‘Huh, okay. So what?’ But in many ways, that inspires the founder to work even harder and to dig more deeply into the problem, because they’re thinking, ‘These people don’t get it! This is actually a really important problem!’”
For Flint, the underlying point is that startup work is personally demanding and requires genuine conviction. If running a startup is advancing an ideology, then the former is like an entry into an art competition with a reproduction of an old master’s work. It is a vitally important effort, but perhaps one that makes for a poor startup. It requires no persuasion, no new belief, and therefore little of the ideological conviction that defines a startup.
But this does not merely make them poor startups — it makes them non-startups entirely. Any organization, sufficiently abstracted, can claim to advance an ideology; a law firm advances justice, a bottled water company (at times) advances the idea that water is not a human right. What distinguishes a startup is that its ideology is its primary concern — that the founder is consumed, above all else, with the question of whether their belief about the world is actually correct. That Berkeley’s entrepreneurial ecosystem rarely asks this question is not an oversight. It is a statement of priorities.
For those who treat entrepreneurship as an intellectual exercise, there are comparatively few resources to know if an argument underpinning a startup embodies the first founder or the latter because the way Berkeley promotes entrepreneurship is by increasing accessibility to opportunity. That’s why our clubs host networking events with powerful VCs and not debates with peers.
That third imbalance suggests an obvious solution to all three problems, and Berkeley already has institutions designed for exactly this purpose. If startups begin as ideological claims about how the world should change, then universities should cultivate environments where those claims can be formed, challenged, and refined.
Yet Berkeley already contains the infrastructure for this kind of intellectual work in its humanities. The university has long been a place where ideas are debated, attacked, and rebuilt in public. Political organizations, publications, and debate forums constantly evaluate arguments not by their popularity or market potential, but by their coherence, originality, and conviction.
Berkeley has arguably always been about marrying technology and the humanities. I believe that the answer to Berkeley’s startup problem is in this marriage. Instead of evaluating startups primarily through metrics such as funding, traction, or market size, universities must also evaluate them as ideological arguments. What belief does the founder hold about the world? What assumption are they challenging? What claim are they making that others consider improbable?
If Berkeley wants to produce more startups, it should stop trying to produce more startups. Instead, it should focus on producing better ideas. None of this diminishes the courage required to build a business. Starting any venture demands risk, persistence, and technical competence, and the businesses produced by Berkeley’s ecosystem deserve continued recognition and support. Berkeley may indeed produce more than 1,600 venture-backed companies. Its entrepreneurial organizations create untold opportunities, opportunities which, as a Berkeley founder, I benefit from. It was through one of these organizations that I met Mr. Flint.
But startups are something else entirely.
People have always found something deeply human in entrepreneurship. I think it’s because vulnerability is fundamentally human. All businesses require the courage to risk one’s finances, career, time, and dedication. But a startup is uniquely human in requiring an intellectual vulnerability. It demands the exposure of a deep-seated worldview and the risk of that highly personal view failing under a spotlight.
Simply put, I wish there were more startups at Berkeley.
Featured Image: Columbia Pictures