Cancer Capital: VC didn’t use to work like this

When I talked about the rise of Cancer Capital, I mentioned that it represents a massive shift from how venture capital has worked over the years. But my conversations in recent years with people in tech, and especially with those outside the industry, reveal that most folks have no idea just how huge that shift has been. It's easy to illustrate exactly how extreme things have gotten just by using a few examples, starting with companies that are familiar to everyone, and sharing some details of what I've seen firsthand.

First: The companies that defined the modern era of tech weren’t founded with venture capital.

Neither Microsoft nor Apple took a penny of venture capital funding when they were founded. Both got started from money they got from their founders and their first customers, and took off from there.

Building for the ages

In Microsoft's case, they didn't get any venture capital investment until the company had been around for six years, and were already doing $17 million dollars a year in revenues. (That was a lot in 1981!) The founders didn't need any startup capital at the beginning because they basically formed the company in order to serve their first customer, and had revenues from the start. More strikingly, Microsoft didn't close their venture capital funding until after they had made their deal with IBM and shipped MS-DOS — the deal that actually made Microsoft into the industry-dominating player that they've been ever since.

And, tellingly, in that funding round, Microsoft only raised $1M from David Marquardt, less than 6% of their annual revenues.

Apple followed a roughly similar pattern. The Apple myth is that they got their initial funding by Steve Jobs selling his VW bus and Woz selling his HP calculator. (Woz told me that story is basically true, not just one of those Silicon Valley narratives that people like to make up.) Their third founder Ron Wayne (the guy everybody forgets about) got too nervous about having to take on liability for the debts that come with starting a new business, and took off. Apple did get some funding in the years that followed — longtime Apple exec, and sometime CEO, Mike Markkula helped fund the Apple II with less than $100k of his own money. And the Apple II was what made Apple the dominant player in personal computers, turning them into the industry force that they've been (except for that near-death moment in the 90s) ever since.

Like Microsoft, Apple didn't take any venture capital funding until January of 1978, when it raised about half a million dollars, nearly two years after it was founded. By then Apple had shipped two products and was already profitable.

These stories were well known and often repeated in the tech industry that I joined at the start of my career. These facts were part of why everyone understood that venture capital was not only not necessary for a company to succeed, but was a resource that was added after a company had already cemented the fundamental strengths that would ensure its success.

Microsoft and Apple had built the businesses and launched the foundational products that would make them legendary before they ever took a single penny of venture capital funding.

The math starts to shift

I'm not telling you these stories out of some nostalgic love for the olden days of computing; there were lots of terrible things about that era, and there have certainly been plenty of harmful actions from both Microsoft and Apple in the years since. I think it's simply essential to reset the framing of what it takes to build a meaningful company at scale. The context of that era matters, because it gives us a useful reference point for everything that follows.

I first learned about venture capital when the startup I was part of in 2003 got its first round of funding. It was for $600,000. Yes, there was a time when an A round of funding was $600,000. What's more, that amount was just about equivalent to what we made in revenue that year. (I'd helped write the business plan, so I knew that we had raised roughly what we earned.) It seems very quaint now, but in the shadow of the dot-com bubble bursting just a short time earlier, even this size of an investment made headlines in the tech trade press.

By the next year, I met David Marquardt myself, as he joined the board of the company and we raised $10 million dollars, at a time when we were probably making at least $4 or $5 million dollars a year in revenue. The key thing here, and why all these numbers matter, is because you can understand the math. It's a fairly straightforward calculation to see what we had to do to get the company to a size where that investment made sense. (And, as a side note: the economics had changed — we were one of the first consumer subscription service products, and one of the very first to run on Amazon Web Services.)

Adjusting for inflation, and accounting for the difference in relative revenues, venture capital had certainly gotten more ambitious in the 23 years since Marquardt cut a check for Microsoft, but it wasn't different in kind. Now, venture-backed companies still often defaulted to compensation structures that were wildly inequitable, and were deeply exclusionary about who was allowed to pitch or get funded. But the core mechanics of how the financial operations were meant to function were possible to discern.

Basically, you could still put all this shit in a spreadsheet.

Where we're at

For a look at the current venture capital landscape, let's consider something like Project Prometheus, which launched in November of 2025. It's an AI startup that's supposed to work on stuff like factories and manufacturing, and it's a Jeff Bezos thing. By contrast to our A round of funding in 2003 of $600,000, Project Prometheus had a seed round of $6,200,000,000. That's not their A round — that's the seed round. Fortunately, that little bit of dipping their toe in the water went well for them, and seven months later, they raised another $12,000,000,000 at a valuation of $41 billion. Not too bad for a company with 150 employees, I hope that scrappy bunch of dreamers can find a way to get by.

And they're not alone. Yann LeCun (he's the AI guy who used to work at Meta, and was supposed to be the nice guy with a conscience here, until surprisingly that clashed with Zuckerberg's priorities) started a company called AMI Labs, and they launched with a billion-dollar seed round, which immediately became the largest in European history. Ineffable Intelligence (AI from ex-Google DeepMind guy David Silver) raised $1.1 billion, without even so much as a published roadmap. And there are lots more. Putting aside these giant headline numbers, just looking at the overall range, the median Series A in America right now is nearly twenty million dollars. When I raised a $30 million A round for Glitch in 2018, it was one of the biggest consumer product A rounds in the entire industry, and was based on both us having had extraordinary user growth and our founders having founded multiple massive companies; just 8 years later, it's below the average for startups overall, including those started by people with no experience and companies with no product or no users.

The difference over 45 years is stark: Microsoft, with a hit operating system and seventeen million dollars a year coming in the door, raised one million. Prometheus, with nothing to show at all, raised six thousand, two hundred times that. Yes, they've got a founder who is enormously rich and famous. But they don't have any products! Or customers! And people hate their founder so much that Katy Perry's album tanked!

Where does that leave us?

But the point isn't the multiples or any complex math. It's simple stuff that a kid with a lemonade stand could understand if they were borrowing money from mom to buy lemons. We raised about what we earned, and then later about twice what we earned, once we had a good idea where things were headed. My biggest mistake in my last startup was over-raising, not seeing how far things had already slipped down this broken path, even though we had a lot of substance to back up what we were doing. And now we're in a place where Prometheus raised $6.2 billion dollars against no discernible product or revenues. There's no reasonable way to value nothing, or to have a ratio against nothing. There's really no way to value a second round of nothing.

But if you're an investor in that follow-on round? Where the valuation increased by billions even though there is no product and no customers? You've already made a massive windfall. You may even have already made enough to turn your entire fund profitable, just on this one deal. Google, which used to employ David Silver, invested in Ineffable Intelligence. And Google will likely be both an investor in its future rounds and potentially the acquirer of the company if and when they ship any products or have any customers, so they will also be able to count the increase in the value of their investment in the company on their bottom line. Cancer Capital has these big companies treating their employees as companies waiting to be spun out, rather than as careers or team leaders they can nurture. And none of that has anything to do with building products, serving customers, taking care of employees, or creating businesses that last.

That’s what I mean when I say this isn’t venture capital anymore. The word “seed” used to describe a stage in a company’s life. Now it describes a relationship between an investor and somebody’s résumé. The language that is being used still consists of the same words that were used half a century ago. It just doesn't mean any of the same things. And it's not how the things around us that have lasted the longest were built.