Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Smack says crypto AI competing on efficiency today is probably not the right wedge; he points to coordination and permissionlessness versus centralized frontier labs. Arthur adds that crypto AI projects enforce encryption and privacy by default while mainstream AI teams often ignore privacy until later.

Related Insights

The initial hype around using crypto for decentralized AI compute has faded due to high costs. Instead, VCs like Dragonfly Capital are focused on agentic payments, where swarms of AI agents will need a global, programmable payment rail for micropayments—a problem blockchain is well-suited to solve.

Because the entire crypto ecosystem is open-source, AI developer tools are exceptionally effective at writing, auditing, and debugging its code. This gives the industry a significant OpEx advantage over traditional finance, whose code is proprietary and not in AI training sets.

The AI community (historically centralized, top-down) and the crypto community (decentralized, bottom-up) have long been at odds. This is changing as AI requires a native financial layer for agents and crypto provides tools to decentralize AI's power, forcing a practical convergence of these two movements.

The convergence of AI and crypto goes far beyond machine-to-machine payments. Haun Ventures is investing in the use of cryptographic tools to solve emerging AI challenges. This includes using blockchains to verify the provenance of AI-generated content and leveraging zero-knowledge proofs to protect user privacy from powerful AI analysis.

AI operates probabilistically, making it great for creative and pattern-matching tasks but unreliable for absolute verification. Crypto is deterministic; its outputs are mathematically guaranteed. This makes crypto the perfect antidote to AI-generated fakes, providing a foundation of verifiable truth that AI cannot replicate.

Centralized AI labs have a massive advantage in capital for compute and data. Crypto offers a coordination layer for decentralized competitors to crowdsource GPUs and data, allowing individual participants to collectively fund and own AI models, creating a viable alternative to the dominance of large corporations.

AI agents are turning to crypto not just for efficiency, but out of necessity. The traditional financial system is a dead end for non-human entities, as an AI cannot get a credit card or open a bank account. Crypto provides the permissionless financial rails required for AI agents to operate and self-replicate economically.

The belief that frontier AI models will capture most of the value is a direct parallel to the failed 'Fat Protocol' thesis in crypto. Instead, value will likely accrue to applications built on top of these increasingly commoditized models, not the infrastructure layer itself.

Sam Dare of Templar frames decentralized AI's mission not as direct competition with giants like OpenAI, but as creating optionality. It enables a new market for those who cannot afford massive, centralized training runs, such as nations seeking "Sovereign AI" or researchers exploring niche pre-training, thereby expanding the market.

For AI agents to be truly autonomous and valuable, they must participate in the economy. Traditional finance is built for humans. Crypto provides the missing infrastructure: internet-native money, a way for AI to have a verifiable identity, and a trustless system for proving provenance, making it the essential economic network for AI.