This is an excerpt from the ninth edition of The Funding sent to our verified subscribers on August 24.
The Funding is a fortnightly newsletter written by Yogita Khatri, The Block’s longest-serving editorial member.
Crypto-AI startups are having their moment. This year, some of the largest fundraises in the crypto space have been driven by ventures with a strong AI connection. Such startups are raking in mega rounds, with some even achieving billion-dollar valuations, all while still in their early stages.
Take PIP Labs, for example. They're the minds behind Story Protocol, a Layer 1 blockchain designed to tokenize intellectual property, ensuring original creators are rewarded in the AI era. Their recent $80 million Series B round, led by a16z crypto, reportedly rocketed their valuation to an impressive $2.25 billion. Then there's Sahara Labs, a crypto-AI startup developing the decentralized platform Sahara AI, which pulled in $37 million in a strategic round co-led by Pantera Capital and Polychain Capital with a promise to fairly compensate AI contributors. And let's not forget Sentient, co-founded by Polygon's Sandeep Nailwal, which grabbed headlines last month with an $85 million mega seed round. Clearly, VCs are placing big bets on the future of crypto-AI.
But what's fueling this surge in interest? To get a clearer picture, I reached out to VCs who have invested in several crypto-AI startups.
"AI is a once-in-a-generation, zero-to-one innovation, especially with the advent of large language models (LLMs)," Jed Breed, founder and general partner of Breed VC, told me. He sees the current excitement around crypto-AI startups as part of a "standard hype cycle," where early enthusiasm and capital flow quickly, even if the actual innovations are still on the horizon.
Rajiv Patel-O'Connor, principal at Framework Ventures, said both AI and crypto are "really big stories" and that combining these two "very optimistic narratives" is driving much of the excitement.
People are waking up to some of the downsides of relying on centralized, closed-source companies to provide AI models, and there's a belief that these downsides could be remedied by decentralized tools, according to Tom Schmidt, general partner at Dragonfly Capital. "We think there's an opportunity for decentralized compute networks to ensure permissionless, ungated access to these models, much in the same way that Bitcoin created the first permissionless, ungated currency," Schmidt told me.
AI isn't just a passing trend; its long-term value is undeniable, according to Erick Zhang, managing partner at Nomad Capital and former Binance executive. As AI and crypto rapidly evolve, their convergence could unlock possibilities once thought unimaginable, drawing even more interest from investors, Zhang told me.
In all, the fusion of AI's transformative potential with crypto's decentralized ethos is seen as a potent mix by investors. But they also voice some concerns.
'Most of these startups will fail'
Dozens of crypto-AI projects have raised funds this year and will continue to raise, but "most of these startups will fail," Zhang told me. He points out that many projects are riding the AI hype without the expertise or a clear value proposition to back it up, chasing trends without a solid plan and "sounding more like sci-fi than reality." This often leads to what he calls "ghost chains and ecosystems," where ambitious promises are left unfulfilled, and resources are quickly depleted. Zhang stresses that while there's a significant amount of capital waiting to be deployed in the crypto-AI space, there's a shortage of legitimate AI-native founders, driving up valuations for early-stage projects.
Schmidt echoes Zhang's concerns but from a different angle. He worries that many crypto-AI projects lean too heavily on decentralization without delivering tangible benefits. "We've seen many times that this is not a recipe for success," Schmidt warns. He believes the real opportunity lies in using crypto to enhance the user experience in ways that centralized AI platforms can't.
Adding to the chorus of caution is Jack Platts, founder of Hypersphere Ventures, who is wary of certain AI applications that may not be relevant to crypto. For instance, he questions whether reinforcement learning and distributed compute for inference really belong in the web3 space. He also notes that AI is currently taking more mindshare and capital from crypto, describing it as the latest "roaring 20s bubble."
Patel-O'Connor of Framework Ventures said achieving product-market fit and building something with real demand that can scale are some of the biggest challenges for crypto-AI startups. Despite these hurdles, he notes that there's a rare influx of talent into the crypto space, particularly in crypto-AI, which he sees as a positive sign for future developments.
Matt Stephenson, research partner at Pantera Capital, underscores the importance of "first principles thinking" in the crypto-AI vertical. He's cautious about what he calls "old wine in new bottles" approaches, where startups repurpose existing ideas without genuine innovation. While experimentation is essential, Stephenson stresses the need for a strong, intuitive sense that a project will work before diving in.
In the end, the crypto-AI landscape is as exciting as it is uncertain. Investors are drawn to the promise of groundbreaking innovations, but they're also keenly aware of the risks. Whether these crypto-AI startups can deliver on their ambitious promises remains to be seen, but for now, they're certainly capturing the attention — and the wallets — of the venture capital world.
