CoinFund Founder: Decentralized AI Counters Government Control

He cited Gensyn and Prime Intellect among teams focused on distributed training. Others include Bagel, Pluralis, Nous Research, Macrocosmos, and Covenant AI. Their work was once dismissed as impossible, he said. Now it shows that distributed training can compete with centralized approaches on cost.

Open source AI models have gained wide adoption but struggle with economic sustainability. No viable business model means open models face difficulty attracting long-term investment. Brukhman acknowledged the gap directly.

Among the teams he cited, only Pluralis has proposed a concrete solution. The approach splits model weights among network participants through a tokenized structure. That creates a financial incentive for contributors while keeping control of the model decentralized. No single entity holds full control, making censorship harder to execute.

Brukhman closed his argument with a question for the broader industry. Will AI become fully centralized under government oversight, or will public, open networks prevail? The answer, he suggested, depends on whether the industry acts on the momentum building in decentralized AI research.