Is Decentralization AI a false issue? Dragonfly partners reveal the true intersection of AI and encryption: agent micropayment.

At a time when the integration of AI and Crypto Assets has become a hot topic, Tom Schmidt, a partner at Dragonfly Capital, expressed a cautious view different from the mainstream during the TBPN Crypto Day program. He pointed out that most current Decentralization AI concepts face significant technical bottlenecks, and instead, stablecoins and Blockchain may find a more practical foothold in applications involving micropayment.

Decentralized AI is hard to implement: Distributed GPUs struggle to handle large models.

Tom Schmidt candidly stated that Dragonfly Capital's view on current projects combining AI with crypto technology is conservative, even bearish.

He pointed out that most projects focus on the concept of decentralized inference (decentralized inference) and GPU resource sharing, but such ideas have appeared in early projects like Golem, yet have never overcome the core challenge: "How to effectively schedule distributed computing power to complete the computation of high-performance AI models?"

If you are thinking about how to practically train a state-of-the-art AI model, you wouldn't use a million consumer-grade GPUs scattered around the world, but rather need a large centralized and high-performance A100 cluster.

Schmidt emphasized: "Computation is not a single-dimensional (monolith)"

The types of resources required for different tasks vary greatly, however, it is difficult for decentralized networks to achieve unified scheduling while simultaneously meeting the demands of "performance" and "stability."

He reminded the market that there is a significant gap between the imagination and reality of "Web3 resource-sharing AI."

(AI should it belong to enterprises or everyone? Psyche partners with Solana to initiate the Decentralization AI training revolution)

The truly feasible intersection: agent micropayments based on stablecoins

Despite skepticism about Decentralization AI, Schmidt does not deny the potential for the combination of AI and encryption technology.

He mentioned that Dragonfly is more interested in applications that combine the deterministic characteristics of Blockchain, especially "AI agents' micropayments (micropayments for AI agents)." In these scenarios, stablecoins can provide programmable and instant settlement capabilities for micropayments, meeting the demands for determinism and finality in many AI applications, which is difficult for traditional financial systems to achieve:

As AI agents become more widespread, the need to pay fees to access data or services is emerging, and they will require a micropayment system capable of handling transactions as low as one cent or even less.

He added, "Traditional SaaS payment providers cannot meet this demand, and 'stablecoins' are uniquely positioned to achieve this."

(Is e-commerce dead? How will AI agency business take over global consumer decision-making power and rewrite business rules?)

Emerging Contrarian Views: Will Decentralized AI Approach from the Perspective of "Attention"?

Another industry figure, Nathan Worsley, founder of Euphoria, starts from the perspective of "attention economy" and believes that AI and encryption technology have a high product-market fit in content distribution and interaction design in Web3 (PMF):

Large language models ( LLM ) and the evolution of AI allow us to analyze posts, videos, and emotions more accurately, depicting a comprehensive outline of "mindshare (" that was impossible to quantify in the past.

)What are you shouting about! A detailed explanation of the widely discussed LOUD mechanism, a new experiment combining attention economy with Ponzi? (

The exploration of AI combined with encryption, moving from speculation to practice.

Tom Schmidt's sober analysis injects a dose of wakefulness into the fiery AI combined with the crypto craze. He cautioned the market not to pin all its hopes on decentralized reasoning that is not yet feasible, but to focus on the application layer where the blockchain can truly exert value. Among them, proxy payments and stablecoin microtransactions may become the path for AI crypto to truly land.

Of course, he is also open to the inclusion of other ideas, injecting more possibilities into this cross-experiment of AI and encryption, collectively driving the evolution of this field.

Is Decentralized AI a False Issue? Dragonfly Partners Reveal the True Intersection of AI and Crypto: Agency Micropayments First Appeared in Chain News ABMedia.