Blockchain AI Criticized for Stifling Decentralized Innovation

Blockchain AI is increasingly being criticized for overshadowing and stifling the development of true decentralized AI, according to an opinion piece by Samuele Marro, a PhD student in machine learning at the University of Oxford [1]. The article highlights how the Web3 AI space has conflated the concepts of “blockchain AI” and “decentralized AI,” leading to a misalignment that hampers innovation.

The author argues that many promising decentralized AI projects are forced to adopt blockchain infrastructure not due to technical necessity, but to access Web3 funding, expertise, and communities. This creates a false equivalence, where blockchain becomes a de facto requirement to be considered a “Web3” project, even when it adds unnecessary complexity, latency, and cost [1].

Web3 ideals—such as trustlessness, permissionlessness, censorship resistance, and user ownership—are distinct from blockchain technology itself. Examples of Web3 include systems like BitTorrent, Tor, and IPFS, none of which rely on blockchain. Similarly, decentralized AI techniques such as federated learning can operate effectively without the need for tokens or on-chain settlement [1].

The article notes that while blockchain has specific applications in AI—such as simplifying payments between agents, enhancing reputation systems, or aligning incentives through tokens—these are niche tools, not universal solutions. The forced integration of blockchain into AI projects is often driven by ecosystem incentives, not product requirements. Venture funds and Web3 communities increasingly exclude non-blockchain AI projects, pressuring teams to conform to prevailing trends rather than pursuing optimal technical solutions [1].

Marro emphasizes that decentralized AI encompasses a range of technologies—distributed computing, P2P networks, edge computing, and more—none of which inherently require blockchain. The conflation of these three distinct concepts—decentralized AI, crypto-integrated AI, and Web3 AI—has led to an artificial bundling that limits creative approaches and real-world applications [1].

To foster true innovation, the author calls for a shift in the ecosystem. Web3 funding and community support must evolve to embrace non-blockchain-based decentralized AI solutions. Examples such as Numerai, Torus Network, and Render Network demonstrate that blockchain can be used effectively when it makes architectural sense, but it should not be a mandatory component [1].

As AI systems grow in power and centralization, the need for decentralized alternatives becomes more urgent. However, this cannot be achieved if the ecosystem continues to impose a blockchain-first approach. The author concludes that Web3 AI must choose between continuing to constrain decentralized AI through blockchain requirements or liberating it to reach its full potential [1].

---

Source:

[1] Blockchain AI cannibalizes decentralized AI

https://cointelegraph.com/news/blockchain-ai-de-ai?utm_source=rss_feed&utm_medium=rss&utm_campaign=rss_partner_inbound