Venture capital is pouring into AI startups with no products or revenue. Founders are using split funding rounds to secure billion-dollar valuations. Investors and employees face new risks as hype outpaces business fundamentals.
In late 2025, David Silver, a former Google DeepMind scientist, pitched his new venture, Ineffable Intelligence, to a group of venture capitalists over Zoom. Silver, known for his work in reinforcement learning, argued that current AI training methods were unsustainable and outlined a vision for AI systems that could learn independently in digital environments. Despite offering no product roadmap or revenue plan, Silver’s reputation drew attention-though some investors left the call unconvinced, describing the pitch as “absurd” and lacking substance.
Yet, Silver’s startup soon made headlines by raising $1.1 billion in seed funding, reportedly Europe’s largest ever, at a $5.1 billion valuation. Company filings reveal a more complex picture: Ineffable Intelligence first secured $11 million from Sequoia and others at a $55 million pre-money valuation. Just weeks later, a second tranche brought in $1.1 billion at a $4 billion pre-money valuation from Lightspeed, Index Ventures, DST Global, and others. This structure allowed early investors like Sequoia to buy in at a steep discount before the valuation soared-sometimes by more than 70 times within a month.
Such split or “tranched” funding rounds are becoming standard in the AI sector, especially among so-called neolabs-startups that raise massive sums upfront to pursue advanced research rather than immediate products. Building these AI labs requires significant capital for GPUs and infrastructure, pushing founders to seek large investments early. In return, lead investors often negotiate better terms for the initial tranche, while later investors pay much higher prices to get in on the action. “At some point there'll be some pushback from VCs, but the market's so hot right now that you have no other choice really if you want to get exposure to some of the best companies,” said Zach DeWitt, partner at Wing VC.
This approach is spreading beyond research labs. Infrastructure provider Baseten recently raised $1.5 billion in two tranches-one at an $11 billion valuation, another at $13 billion. Other AI startups, including Aaru and Serval, have reportedly used similar tactics, according to sources cited by Forbes. “It’s very common among valuation maximizing founders,” one venture capitalist noted.
For founders, these headline-grabbing valuations can attract top talent and future investors, even if the underlying business is still unproven. Employees, however, may be misled by the publicized numbers, as their equity is often priced closer to the higher valuation, not the blended average of both tranches. “They’re taking more of the risk and capturing less of the upside,” said Jaya Gupta, partner at Foundation Capital. “The social contract that made startup equity compelling is eroding and most candidates don't realize it until the exit happens (or doesn't).”
Investors also use these rounds to signal credibility. The lead investor in the first tranche often helps coordinate the entire round, drawing in strategic players like Nvidia, Google, or Microsoft, who are less sensitive to price and may see their investments convert into lucrative chip or cloud contracts. Blue-chip VCs now encourage startups to raise both tranches simultaneously, leveraging their brand to create urgency and competition among later investors.
Some industry figures have criticized the practice. Brendan Foody, CEO of Mercor, called it the “Sequoia Scam” in a June post on X, alleging that founders and investors often misrepresent the true terms to employees and angels. Sequoia partner Shaun Maguire responded that the firm has used this structure only a handful of times, typically when other investors are willing to pay much higher prices for hot AI startups.
For VCs, the upside is clear: early entry at a low price, followed by a rapid markup as later investors buy in at inflated valuations. Menlo Ventures partner Deedy Das estimated that over 63 neolabs have collectively raised about $48 billion and are valued at more than $300 billion, representing a significant share of recent AI investment outside of giants like OpenAI and Anthropic.
Government-backed funds are also joining the frenzy. The United Kingdom’s Sovereign AI fund and the British Business Bank, both supported by taxpayer money, invested in Ineffable Intelligence, though their entry valuation remains undisclosed. This influx of public and private capital is reshaping the competitive landscape for AI startups.
As the funding race accelerates, some investors worry about the long-term consequences for employees and the sustainability of these valuations. The risks are especially acute for those joining after the funding rounds, as they may pay a premium for equity that could take years to realize-if ever. For founders like Silver, the gamble is clear: “There will be a significant risk of failure for a chance of spectacular success that will positively transform the course of AI and thereby humanity,” reads a January blog post on Ineffable Intelligence’s website.
This trend echoes broader shifts in AI startup funding, where distribution strategies and investor hype can outweigh technical milestones. For more on how go-to-market tactics are redefining AI startup success, see this analysis of the changing competitive edge in the AI sector.
According to public filings, Ineffable Intelligence was founded in 2025 and quickly became one of Europe’s most heavily funded AI startups. The company’s $1.1 billion seed round, split across two tranches, drew participation from Sequoia, Lightspeed, Index Ventures, DST Global, and government-backed funds. As of mid-2026, Ineffable Intelligence has not yet launched a commercial product or reported any revenue.
