"Too Much Capital, Too Few Deals": The Great Venture Capital Lie
- Aki Kakko
- 2 minutes ago
- 6 min read
If you sit at a dinner table with General Partners from top-tier venture firms, you will inevitably hear a variation of the same complaint:
“The Venture Capital industry is over-saturated. There is too much capital chasing too few deals.”
This sentiment has become the accepted wisdom of the post-Zero Interest Rate Policy (ZIRP) era. It posits that the number of Power Law winners—the next Googles, Ubers, and OpenAIs—is a fixed variable. Therefore, pouring more money into the ecosystem doesn’t create more innovation; it simply inflates the entry price of the few "obvious" winners. According to this worldview, the economy is a parking lot with a finite number of spaces. If you add more cars (capital), you don’t get more parking; you just get traffic.
This is a fundamental misunderstanding of how innovation works.
The "Inelastic Supply" argument is a static view of a dynamic system. It looks at the economy as it is today and says, "There is no room." But innovation is not about parking in existing spaces; it is about expansion and building new floors.
The number of winners is not fixed by the market. It is limited only by the imagination of founders and the efficiency of the capital available to fuel them.
The Static Fallacy: Why VCs See a "Fixed Pie"
The "Too Few Deals" argument is based on a zero-sum economic model. It assumes that categories are finite containers.
"Search is won."
"Social is won."
"Cloud is won."
"LLM is won."
If you believe the pie is fixed, then yes, Venture Capital is merely an auction house where 500 firms bid up the price of the same three AI startups.
But this ignores the Fractal Nature of Innovation.
Technology is not a game of musical chairs; it is a game of platform creation. Every time a major "category winner" is established, it creates surface area for thousands of new companies to be built on top of it.
The Platform Effect: When Apple "won" the smartphone category, they didn't close the market. They created the App Store, which birthed Uber, Instagram, WhatsApp, and Mobile Gaming—entire industries that could not exist before the platform was established.
The Unbundling Effect: In 2000, VCs thought the "Classifieds" category was finished because of Craigslist. Then Airbnb unbundled the "Housing" tab. Tinder unbundled the "Personals" tab. Indeed unbundled the "Jobs" tab.
The Evolution Effect: VCs today act as if the Transformer architecture (LLMs) is the "end of history" for AI. This is akin to believing the Vacuum Tube was the peak of computing. LLMs are powerful, but they are fundamentally probabilistic token predictors. The next wave of Power Law winners will not come from "better chatbots," but from entirely new architectures—neuro-symbolic systems, active inference models, or "System 2" reasoning engines—that solve the fundamental limitations of energy efficiency, causality, and hallucination that LLMs cannot. We are not at the end of the AI roadmap; we are barely out of the driveway.
The Compute Cycle: The market currently acts as if the "Chip War" is over because Nvidia is dominating the current LLM training phase. This ignores the history of hardware. Intel’s dominance in CPUs did not prevent the rise of ARM in mobile or Nvidia in graphics. We are now shifting from the era of Training (massive, energy-hungry clusters) to the era of Inference (running models everywhere). The GPU is not the final form of compute; it is simply the brute-force engine required to bootstrap the current LLM based intelligence. The next generation of hardware giants will be built on entirely new physics—optical computing, neuromorphic chips, and specialized ASICs—that prioritize energy efficiency and edge latency over raw parallel processing.
The economy is a positive-sum expansion of the frontier. Every time a "hard" problem is solved, it lowers the barrier to entry for solving the next ten problems.
The Streetlight Effect
If the supply of opportunities is actually expanding, why do smart investors feel like it is shrinking?
Because of the Streetlight Effect.
A policeman sees a drunk man searching for his keys under a streetlight."Did you lose them here?" the policeman asks."No," the drunk man replies, "I lost them in the park, but this is where the light is."
Traditional VCs are huddled under the same streetlight—SaaS, Consumer, and Generative AI—in the same three geographies (SF, NY, London).
Because the Artisanal VC Model relies on warm intros and consensus pattern matching, every fund sees the exact same deal flow.
They are fishing in the same pond with 500 lines in the water, complaining that there are no fish, while the ocean behind them is teeming with life.
The scarcity is not in the deals; the scarcity is in the perspective of the investors.
Jevons Paradox: The Infinite Supply of Problems
The "Inelastic Supply" argument also assumes that demand for solutions is finite. But the Jevons Paradox teaches us that as technology increases efficiency, consumption rises. As AI, cloud computing, and no-code tools drive the cost of starting a company toward zero, the definition of a "venture-backable problem" expands radically.
We are moving from a world of "Tens of Giant Companies" to a world of "Millions of Specialized Agents." The supply of human problems to be solved is effectively infinite. The only bottleneck is the mechanism to fund them.
This disconnect between the potential supply of companies and the actual capital available creates an Innovation Deficit. Thousands of viable, high-growth businesses die every year. They do not die because they lack merit. They die because:
They are in "boring" industries (Logistics, Agriculture, Manufacturing).
They are in "wrong" geographies (Jakarta, Lagos, Helsinki, Ohio).
They do not fit the "Pattern Match" of the 1000 people who control the capital.
When VCs say "There are too few deals," what they really mean is:
"There are too few deals that fit my narrow thesis, my warm intro network, and my 10-year fund constraints."
Closing the Deficit: The Self-Service VC Model
To unlock the "Infinite Frontier" of innovation, we cannot rely on the boutique, artisanal model of 20th-century Venture Capital. We need a capital market that scales as easily as the cloud. We need Self-Service VC.
Self-Service VC is the infrastructure required to move from an economy of Scarcity to an economy of Abundance.
Meritocracy over Network: By allowing founders to connect data directly, we remove the "Warm Intro" filter. We judge the business on its physics (traction, retention, growth), not its metaphysics (founder pedigree, narrative).
Volume as a Strategy: Instead of trying to pick 10 winners, we can fund 100,000 experiments. We acknowledge that we cannot predict the Power Law ex-ante, so we maximize our exposure to the ecosystem.
The Efficient Frontier: By automating the legal, operational, and diligence layers, we can profitably write smaller checks into earlier companies. This allows us to capture the "Long Tail" of innovation that the artisanal model finds "too expensive to diligence."
Fueling the Bakery
The belief in "Too Much Capital, Too Few Deals" is a self-fulfilling prophecy. If you build a financial system that only has the bandwidth to process 100 companies, the market will warp to look like there are only 100 good companies. But if you look at the world through the lens of history, you see that human ingenuity is boundless. We have never run out of problems to solve. We have only run out of the capital efficiency required to solve them.
Self-Service VC is not just a different way to invest; it is a mechanism to close the Innovation Deficit. By aligning capital with the true dynamic nature of the economy, we stop fighting over a fixed slice of the pie and start fueling the bakery.
The "Power Law" Trilogy:
Second chapter: Why the Smartest Money in VC Looks the Stupidest
