Beyond Central Control: What is Decentralised AI?
For most of us, AI lives in the cloud, on servers owned by massive corporations. When you use a tool like ChatGPT, you are accessing a centralised model. Decentralised AI flips this script. It combines artificial intelligence with technologies like blockchain
to distribute computing power, data, and even the AI models themselves across a peer-to-peer network. Instead of one company controlling everything, a global network of independent nodes contributes resources, processes information, and maintains the system. Think of it as the difference between a single, powerful radio station broadcasting to everyone versus a network of community radio stations, each serving local needs while sharing a common framework. This approach moves away from a monopolised structure towards open, community-governed models that lack a single central owner.
Altman's Vision: Democratising AI Power
In recent interviews and statements, Sam Altman has expressed fears that AI technology could become concentrated in the hands of too few companies, leaving society without a say in how it evolves. His vision is for people to have deep control over AI's future, allowing society and the technology to co-evolve. This push for 'democratisation' aims to oppose the concentration of power. Altman argues that beyond simply giving everyone access, key decisions about the technology should be made through democratic processes, not just inside AI labs. This philosophy also underpins his evolving thoughts on economic fairness. He has shifted from supporting Universal Basic Income (UBI) to ideas like Universal Basic Compute, where individuals might receive a share of AI computational power that they can use, sell, or donate, giving them a direct stake in the new economy.
An Economic Boom for India's Tech Sector
For a nation with a thriving technology and startup ecosystem like India, the implications of decentralised AI are profound. Altman himself has predicted that AI will empower the "greatest boom in people starting smaller businesses than we have ever seen". A decentralised model could dramatically lower the barrier to entry for Indian developers and entrepreneurs. Instead of relying on expensive, proprietary models from Silicon Valley, they could tap into open networks, customise models for local languages and contexts, and build solutions for uniquely Indian problems. This fosters innovation and promotes data sovereignty, ensuring that sensitive Indian data can be managed within the country. The ability to build and deploy AI without permission from a central gatekeeper could unleash a wave of grassroots innovation in sectors from healthcare to finance and agriculture, creating economic value across the board.
The Hurdles on the Road to Decentralisation
Despite the promise, the path to a decentralised AI future is filled with significant challenges. Centralised systems offer consistency and are highly optimised, while decentralised networks can be complex to manage and may face scalability issues. Ensuring quality control, security against malicious actors, and consistent performance across a distributed network of nodes requires advanced tools that are still evolving. There are also regulatory and legal concerns to navigate. Furthermore, Altman has conceded that the economic and societal adaptation to AI is happening more slowly than he initially predicted. Businesses need time to change processes, manage budgets, and understand new capabilities before they can fully adopt and integrate these new technologies. This lag, however, might be a good thing, allowing for a smoother, more stable transition.
