ai-behind-closed-doors-centralization-may-become-threat

AI behind closed doors: why centralization may become a threat

AI

January 15, 2025

Modern artificial intelligence models, such as OpenAI, are increasingly criticized for their lack of transparency. Simon Kim, CEO of the crypto venture fund Hashed, called this phenomenon "creating a god we don't understand." Closed models operate in an environment of strict centralization, where access to technology and control remain in the hands of a few. According to the expert, this poses a threat to public interests and raises concerns about potential risks.

The key to transparency

Kim believes decentralization and transparency should be the foundation of AI development. Blockchain could play a key role by providing tools to track content usage and ensure fair compensation for creators. It would also make generative models transparent, accessible for analysis, and subject to control. Today, AI uses data without compensating the creators, but decentralized technologies could create a fair ecosystem for all participants.

First steps toward a new AI era

Open-source tools like Meta's Llama reveal the potential of decentralized approaches. Kim suggests going further by creating an "author layer"—a platform to track content usage and reward copyright holders. Examples include projects like Story or Frax Finance's developments for AI agents on the blockchain.

Not everyone shares the enthusiasm about integrating blockchain and artificial intelligence. Some venture capitalists express skepticism, believing that the practical implementation of such projects is still far from perfect. Nevertheless, industry leaders, such as former Binance CEO Changpeng Zhao, consider this alliance promising.