What Venture Capital Calls Innovation.

Introduction

There is a word that the venture capital industry has borrowed from the people who actually did the hard thing. The word is “innovation,” and the borrowing happened so gradually, and so completely, that most people now use it to mean its opposite.

The original meaning was not complicated. Innovation meant the production of something genuinely new: at real cost, over uncertain timescales, by institutions with no certainty that anyone would want what they were building. That description fits exactly one set of institutions that have been operating in the United States and Europe for the past seventy years. It does not describe venture capital.

What the Record Shows

The internet began as ARPANET. In December 1968, the United States Department of Defense contracted Bolt, Beranek and Newman, a Cambridge research firm, to build the first routers for a sum of approximately one million dollars. The first message was transmitted on October 29, 1969, between UCLA and the Stanford Research Institute. The network crashed before the message finished. No investor was waiting for a return. No founder held equity. A defense agency had decided that resilient, distributed communications were worth building, and had found the money.

Over the following twenty-five years, DARPA-funded researchers expanded the network and developed TCP/IP, the protocols that allow computers to identify and reach each other across it. By the time commercial internet companies began receiving venture funding in the mid-1990s, the underlying infrastructure was already built, already proven, and open to anyone who wanted to build on it. The investors who funded those companies did not fund the internet. They funded applications on top of something the public had already paid for.

GPS follows the same line. The Department of Defense approved its development in December 1973. By 1995, when the full constellation of twenty-four satellites became operational, the total cost to the public was estimated at approximately $12 to 14 billion. Reagan opened civilian access in 1983, following the Korean Air disaster. Clinton granted full civilian access to ungraded signals in 2000. Every navigation application, every rideshare dispatch system, every logistics platform that a venture firm has ever described as an innovation runs on a satellite network the American public paid for across two decades before a single commercial application existed.

The World Wide Web was invented in 1989 by Tim Berners-Lee, then working at CERN, the European Organization for Nuclear Research, funded by its European member states. He was solving an information-sharing problem for physicists. His supervisor wrote “vague but exciting” on the initial proposal. In April 1993, CERN released the technology as public domain, free to the world. It was not sold. It was not licensed. A publicly funded physics laboratory handed the architectural layer of the modern commercial economy to anyone who wanted it, at no charge, and the industry built on top of it now describes itself as the primary engine of innovation.

Where the Venture Money Actually Entered

Two cases allow the timing to be seen precisely, because the documentation is unusually complete.

Larry Page and Sergey Brin developed the PageRank algorithm as Stanford graduate students working under the NSF-funded Stanford Digital Library Project, supported by National Science Foundation Cooperative Agreement IRI-9411306, a grant of approximately four and a half million dollars. Brin held an NSF Graduate Research Fellowship. The 1998 paper that introduced Google and PageRank to the world closed with an acknowledgments section that named NSF, DARPA, and NASA as its funding sources. Sequoia Capital and Kleiner Perkins arrived with $25 million in June 1999, nearly three years after the breakthrough, after the algorithm was proven, after the academic papers establishing its significance had been published, and after early users were already relying on it inside Stanford’s network. The public money absorbed the foundational risk. The private money arrived once that risk was substantially gone.

The mRNA timeline is longer, and the pattern is more instructive. Katalin Karikó submitted her first NIH grant proposal to use messenger RNA as a therapeutic in 1989, when she joined the University of Pennsylvania. Funding was difficult from the beginning. Penn demoted her in 1995 for failing to secure sufficient grants. She continued the work at reduced standing. In 1997, she met immunologist Drew Weissman, who had trained under Anthony Fauci at the NIH, at a photocopier in a Penn laboratory. Their 2005 paper on modified nucleosides, the discovery that made mRNA stable enough to use therapeutically, was rejected by Nature and then by Science before Immunity published it. BioNTech hired Karikó in 2013, twenty-four years into her publicly funded research, to apply what she had already built. The foundational patent covering the nucleoside modification that made mRNA therapeutically viable was funded by the NIH. The pharmaceutical companies that commercialized the work are regularly described, in industry communications and in general coverage, as the innovators. Karikó and Weissman received the Nobel Prize in Physiology or Medicine in 2023, for research that had been largely invisible to private capital for three decades.

What Venture Capital Actually Did

None of this is an argument that venture capital should not exist, or a claim that its work lacks value. Getting the internet into the hands of ordinary people required product development and deployment capital at a scale that government agencies had neither the mandate nor the mechanism to provide. Sequoia’s investment in Google in 1999 required genuine judgment about a company whose revenue model did not yet exist. The commercialization of mRNA technology into a vaccine that reached billions of people in under a year represents a real achievement in execution. The claim here is narrower than an indictment.

The claim is about a word.

When the venture capital industry describes itself as an innovation engine, the word is covering the distance between where the technology came from and who is now receiving credit for it. What venture capital actually does, accurately described, is identify technologies that have already been substantially derisked, frequently by public institutions working on fifteen-to-thirty-year timescales, and fund their commercial scaling at a stage when a return is plausible. That is an honest description of a real and often valuable activity. It is not a description of the thing that produced the internet, or GPS, or the World Wide Web, or the PageRank algorithm, or the mechanism inside every mRNA vaccine.

The structural reason this gap exists is not a failure of intention. A standard venture fund has a ten-year life. General partners typically raise new funds every two to three years, which creates pressure to show returns well inside that window. The research that produced the commercial internet ran from 1969 to the mid-1990s, approximately twenty-five years. Karikó’s work ran thirty. No fund structure in venture capital is built to absorb foundational risk on those timescales, and no limited partner would wait. The institutions that took the foundational risks were not making a choice that private capital could have made instead. They were operating in a category that the structure of private capital cannot reach.

What the Word Is Protecting

The word “innovation,” applied to venture capital, protects a particular version of events. In that version, the private sector takes the risks that produce civilization-level change, and the returns that flow to it follow as a matter of logic rather than a matter of timing. The record does not support this version. The record shows a different sequence: public institutions absorb the foundational risk, over timescales that no fund can match, and release the results into a common pool. Private capital then identifies the most promising applications within that released infrastructure, scales them effectively, and receives both the financial return and the reputational credit for the underlying discovery.

In 2013, economist Mariana Mazzucato documented this pattern in detail across multiple industries, including an analysis of how the core technologies inside the iPhone traced back to DARPA, the Department of Defense, and the Department of Energy rather than to private capital. The pattern she described has not changed in the years since she described it.

Conclusion

The word “innovation” did not migrate through any single deliberate act. The industry that used it most loudly also narrated its own story most loudly, and after a certain point the origin of the word no longer traveled with the word itself. What remains is a vocabulary built to describe one set of activities while pointing at another.

The grant numbers are in the acknowledgment sections of the original research papers. The 1998 paper that introduced Google named NSF, DARPA, and NASA as its funders, in print. The Department of Defense publishes its budgets. Nothing in the record has been concealed. The industry that arrived afterward simply told its own story more loudly, and that story is now the one most people have heard.