Ignore the headlines. Watch the capital flows.
When Lightspeed Venture Partners—a firm with a reputation for early-stage precision—raises a $600 million secondary fund to buy used shares of OpenAI and Anthropic, it's not a bullish signal. It's a liquidity admission. The same mechanics that drove crypto into a bear market in 2022 are now surfacing in AI: overvalued assets, locked-up equity, and a desperate need for exit liquidity.
Context
Lightspeed's plan, reported by Bloomberg and filtered through Web3 media, is straightforward: raise a dedicated vehicle—codenamed Project Mercury—to purchase existing shares from early investors and employees of the two largest AI labs. The fund is structured across multiple vehicles: Select V, Opportunity II, and separate managed accounts. The target is $600 million, but the real story is not the size—it's the structure.
Secondary funds are the financial equivalent of a distress signal. They exist when primary markets can't absorb supply. In crypto, we saw this in 2018 when ICO tokens traded at 90% discounts on secondary markets. In 2022, it was the collapse of centralized lending platforms. Now, it's AI's turn. The fact that Lightspeed—a firm that minted its reputation on early-stage bets—is now buying secondhand shares tells you that the primary market for AI equity is clogged.
Core Analysis: The Dual Hedge and the Liquidity Trap
Lightspeed is not betting on a winner. It is hedging against the possibility that both lose.
By simultaneously adding exposure to OpenAI and Anthropic, Lightspeed acknowledges a truth that most AI evangelists ignore: the market cannot determine a winner. This is not a contrarian take—it's a structural admission. In 2017, I audited 12 ICO whitepapers, including EOS and Tezos. I saw the same pattern: teams raising massive funds with no clear path to liquidity. The smart money diversified. The smart money hedged. Lightspeed is doing the same.
Bets are cheap; exits are expensive.
OpenAI's annualized revenue is rumored to be in the billions, but its training costs outpace revenue. Anthropic's Claude has carved a niche in enterprise safety, but its growth is measured in months, not quarters. Both companies are burning cash at rates that would make a crypto DeFi yield farmer blush. The secondary fund provides a release valve for early employees and funds that need to cash out. But who is the buyer? Lightspeed is effectively acting as a market maker, offering liquidity at a discount. The price of that discount is the implicit cost of the market's inability to price these assets.
The $600 million figure is a red herring.
In the context of AI's multi-billion-dollar funding rounds, $600 million is not a bet on the future—it's a parking lot for capital that cannot find higher-yielding opportunities. Lightspeed's LP base likely includes sovereign wealth funds and pension funds that are seeking long-term exposure without the risk of primary allocations. This is capital that would have flowed into crypto three years ago. Now, it's chasing AI. The shift is a macro signal: the liquidity that once fueled Web3 is now being absorbed by AI's infrastructure narrative.
Contrarian Angle: The Secondary Market Is the Canary
The common narrative is that AI is booming and VCs are doubling down. The contrarian view is that the secondary market is a symptom of a market that cannot sustain its own valuation. In 2020, I managed a $15 million DeFi portfolio. I saw the same pattern when Curve Finance's staking yields dropped: liquidity providers exited, and the protocol's TVL collapsed. The same principle applies here. When the primary market cannot provide exits, the secondary market becomes the only game in town. And secondary markets are always priced at a discount to the last round.
The dual investment in both OpenAI and Anthropic is not a strategy—it's a surrender.
It says that even the most informed capital cannot predict which model will dominate. This is the same logic that drove investors to buy both Bitcoin and Ethereum in 2017. It worked then, but only because the market expanded. In AI, the expansion is limited by enterprise adoption rates and regulatory headwinds. If the market contracts, both bets lose. Lightspeed is not hedging against a winner-take-all outcome; it's hedging against a scenario where no one wins soon enough.
Follow the gas, not the hype.
In crypto, "gas" refers to the cost of computation on Ethereum. In AI, the gas is the cost of training and inference. Lightspeed's secondary fund is a bet that the gas will keep flowing—that OpenAI and Anthropic will continue to raise capital at higher valuations. But the secondary market is a leak in the pipeline. Every time an early investor sells, it signals that the confidence in the next round is lower than the confidence in the current price. That is a bearish signal, not a bullish one.
Takeaway: What This Means for Crypto
Lightspeed's move is a microcosm of a larger trend: capital is rotating out of speculative assets and into infrastructure. AI is the new infrastructure. But the same dynamics that made crypto a liquidity trap are now infecting AI. The secondary fund is a canary in the coal mine. If Lightspeed's fund closes below $600 million, the AI party is over. If it oversubscribes, the rot has only just begun.
As a crypto fund manager who has survived three bear markets, I know one thing: liquidity is the only thing that matters. Watch the flows. Ignore the narratives. The moment a secondary fund appears, the primary market is broken. Lightspeed is not building the future—it's buying time. And time is the most expensive asset in any market.