Venture capitalists are increasingly placing large bets on emerging artificial intelligence (AI) startups, commonly referred to as "neolabs," despite many lacking established products, markets, or revenue streams. Over the past two quarters, neolabs have collectively raised $24 billion, an amount nearly five times greater than what leading AI firms OpenAI and Anthropic secured before the introduction of ChatGPT, according to analysis by Radical Ventures.

These new entrants often feature high-profile founders and boast significant valuations even at early stages. For example, Emulate, founded by former Google DeepMind researchers and only a month old, is on track for a nearly $4 billion post-money valuation. Similarly, Safe Superintelligence, launched in 2024 by OpenAI co-founder Ilya Sutskever and yet to unveil a tangible product, achieved a valuation of $32 billion during its recent funding round.

The foundation of these neolabs relies heavily on two critical resources: access to skilled research scientists and cutting-edge computing infrastructure, including chips and servers. Both are limited and costly, contributing to the substantial capital consumption typical of these ventures. Fundraising rounds occur more frequently than in traditional sectors, with an average interval of about 7.6 months, and valuations often increase by a factor of 3.5 between rounds, based on data from venture capital firm Chapter One.

One notable example is Reflection AI, which develops customizable open-weight AI models. The company raised $105 million at a $555 million valuation initially, followed by a subsequent funding round barely a year later that valued it at $25 billion—more than 25 times the original figure. Such steep valuation growth is sometimes facilitated by investors entering quietly at lower valuations before participating in more publicized, higher-valuation rounds. This approach allows founders to promote larger valuation figures, while investors can average out their entry costs.

While the investment strategy carries inherent risks comparable to other venture capital endeavors, neolabs differ from traditional startups in sectors like retail or ride-sharing due to the rapidly evolving nature of AI technology and business adoption. Some neolabs specialize in niche areas such as "world models" that simulate complex environments or "physical AI" that underpins autonomous machines, positioning themselves to capture emerging opportunities.

The AI market remains highly dynamic, with customer preferences and uses still in flux. Even large established companies are frequently exploring multiple AI tools and offerings, intensifying competition among providers. Although it is currently uncertain which neolabs will ultimately establish themselves among the industry’s leading players, investors who diversify their portfolios across several hopeful startups seem to be adopting a pragmatic approach amid this evolving landscape.