Crypto VC: AI crypto projects are unlikely to disrupt major centralized players this cycle
While the potential for projects at the intersection of crypto and AI has driven massive investment from retail and VC investors, two prominent venture capitalists expressed skepticism about the likelihood of crypto projects disrupting major AI players this cycle and implored builders to “follow the incentives.”
“Crypto has a problem of making things that already exist well off-chain bad and then putting them on-chain and selling them to people who are already on-chain, and we see this in all narratives across all asset classes… A lot of what’s happening at the AI/crypto intersection is essentially that,” said Rob Hadick, general partner at Dragonfly, in a statement.
The most likely value AI will bring to crypto is simply helping developers code more efficiently, as software development remains a core element of the crypto space, panelists noted at The Block’s inaugural Emergence conference in Prague on Friday. “AI is a massive software technology revolution,” said David Pakman, managing director of CoinFund. “We are building a lot of software on Web3, and we should benefit from the use of AI in a variety of different decentralized applications, both in terms of infrastructure, security, and how user applications behave.”
However, when it comes to challenging the few large centralized players in the AI space, including OpenAI, Anthropic, Microsoft, and Facebook, the two venture capitalists largely agree that decentralized AI systems are unlikely to surpass these centralized players this cycle.
Hadick further said, “We (Dragonfly) want to see a little bit of product-market fit…We’ve always been skeptical of the idea that there’s going to be a market around things like decentralized training…Doing anything around on-chain models or on-chain ML, we’ve made a few investments…One of them is specifically in edge computing, think about local models, they’re specifically for latency-insensitive workloads.”
Pakman agreed that decentralized projects are unlikely to disrupt centralized players in this cycle, but he praised builders who are trying to develop alternative systems, “If we want a highly centralized AI future, we should continue on the path we’re on now, but if we want an alternative future where the compute resources for training really large AI models are more decentralized and maybe democratically allocated access, then we should talk about what Web3 is doing well in decentralized technology stacks or hardware access.”
In addition, Hadick pointed to proxy-based projects as an area he is particularly skeptical of, likening the proxy hype to crypto blockchain gaming in 2021. He said: "We have an investment in AI agents... This could be a winner-take-all market. I expect tens of millions or even hundreds of millions of dollars in crypto venture capital to enter the space, but I think almost everyone will lose a lot of money." Pakman disagreed slightly, elaborating on how an agent-based future could lead to synergies between AI and crypto.