Earlier this year, David Silver, a distinguished former researcher at Google DeepMind, participated in a Zoom pitch for his new venture, Ineffable Intelligence, aimed at attracting venture capital. With over ten years at Google under his belt, Silver criticized existing AI training methods as unsustainable. He shared his vision of developing digital environments for AI systems to learn independently without relying on human-sourced data.
During the call, he expressed a bold prediction: AI would eventually be capable of interacting with the physical world, even hinting, “We can put AI in our toasters,” as recalled by one investor.
However, many found this concept hard to digest. An investor who attended the pitch said, “I was so excited to hear it, but it was quite absurd. He went on without any visual aids or a structured plan.” By the end of the presentation, the investors felt unenthusiastic. “We walked away with more questions than clarity, which is surprising coming from someone with so much credibility in the field,” they remarked.
Despite lacking immediate product plans or revenue strategies, Silver managed to secure $1.1 billion in seed funding for Ineffable Intelligence. This funding round was celebrated as Europe’s largest seed round, giving the young startup a staggering valuation of $5.1 billion—according to mainstream reports.
In actuality, the funding was executed in two phases. Initially, Ineffable gathered $11 million from Sequoia and other investors, resulting in a pre-money valuation of approximately $55 million. Shortly after, another round raised $1.1 billion with a much higher pre-money valuation of $4 billion through investments from prominent firms like Lightspeed, Index Ventures, DST Global, and Sequoia at the inflated price. This represented a valuation increase of over 70 times within a matter of weeks.
A significant portion of Sequoia’s investment occurred at this elevated valuation, although both Sequoia and Ineffable declined to comment on the matter.
Such tiered funding arrangements have become increasingly common in the current AI funding landscape, especially among new research labs—known as neolabs—raising substantial capital at the outset to fuel cutting-edge research instead of immediate product development. Establishing an AI lab necessitates vast amounts of GPU resources, which requires an influx of funding right away. Investors are demanding more favorable terms due to the large sums involved, according to Zach DeWitt, a partner at Wing VC.
“As the market is so competitive, there’s little choice for VCs if they want exposure to leading startups,” he explained.
This pattern of segmented funding has also emerged in various AI businesses beyond just research labs. For instance, Baseten, a company that supplies AI infrastructure and services, recently completed a $1.5 billion funding round across two phases—one at an $11 billion valuation and the subsequent at $13 billion. Other notable AI startups, such as Aaru and Serval, have also pursued similar funding strategies, as multiple sources have confirmed. “This approach is popular among founders aiming to maximize their valuations,” remarked one venture capitalist.
In a fundraising environment driven by public perception, flashy billion-dollar headlines often hold more value than accurate assessments, Jaya Gupta, a partner at Foundation Capital, noted.
The actual value of such startups is typically represented by a blended valuation—a weighted average reflecting the different valuations associated with each funding tranche, which often goes unreported. Since the specifics of these funding structures are rarely disclosed, larger headline numbers can mislead observers into believing influential firms like Sequoia have an unwavering confidence in the startup’s worth.
Brendan Foody, CEO of Mercor, noted on social media earlier this year that he had observed multiple funding rounds where Sequoia opted for two tranches, dubbing it the “Sequoia Scam.” However, he later clarified that such practices are standard across the industry among top firms.
In response to Foody’s claims, Sequoia partner Shaun Maguire insisted it was unjust to label it a scam, arguing that it was not an uncommon scenario during his tenure. “Other investors often are willing to pay a premium for in-demand companies,” he explained.
This layered funding strategy allows firms like Sequoia to invest in top-tier teams at a lower price, thereby securing more equity and enhancing their potential returns while capitalizing on promising concepts.
The financial stakes are significant; another Menlo Ventures partner, Deedy Das, reported in May that over 63 neolabs have an aggregate valuation exceeding $300 billion, raising around $48 billion. Although it’s unclear how many of these lehabs utilized segment funding, they account for a substantial portion of the total venture investments made outside established entities like OpenAI or Anthropic over the past year.
Venture capitalists orchestrating initial funding rounds also attract strategic investors like Nvidia, Google, or Microsoft, enhancing the startup's perceived potential. These corporate investors often show increased willingness to engage financially since large sums invested can lead to future contracts for chips and cloud services. Prominent venture capitalists are now encouraging simultaneous larger second round funding as a tactic to generate interest and competition among investors.
The disparity in pricing between two VC firms for the same startup arises because those who missed early opportunities with major AI players—like OpenAI—view neolabs as viable options to place their bets on the next wave of transformative startups. There’s also the enormous possibility of profitability, as backing a startup could result in massive gains akin to those achieved with leading firms.
The influx of funds is not limited to Silicon Valley alone; Silver’s Ineffable Intelligence has gained backing from the UK’s Sovereign AI fund and the British Business Bank, with resources rooted in taxpayer money. However, details regarding the valuation at which they invested remain undisclosed since these institutions did not respond to inquiries.
The implications for employees considering these valuations as they choose which startups to join may be significant. New hires might find they are paying a premium for stock options that correspond to the inflated valuation rather than the blended valuations reflecting previous investments. “They are assuming more risk while capturing less potential upside,” Gupta highlighted, pointing out the diminishing value of startup equity incentives, often unrecognized until the company’s eventual outcome becomes apparent.
For companies like Ineffable Intelligence, utilizing this funding structure makes strategic sense, as it allows them to choose preferred investors while attracting large sums necessary for their ambitious research endeavors. As Silver himself alluded, the journey might be fraught with challenges, but the potential rewards for success in AI could be monumental.
This report has been updated to provide further clarification regarding Sequoia's investment in Ineffable Intelligence.
