In the past couple of years, the general public has gotten glimpses into the possibilities of AI, but mostly through the strict bottleneck of large corporations. These firms control how, and when users engage with AI. This is often portrayed in the media as something to pull away from - something negative. There’s a sense that as the tech giants' control grows, the agency of the end user shrinks in ways that can’t be anticipated and all of us are worse off for it.
In truth, what Big Tech does offer is something that can only be delivered by a centralised power - that is, Big Tech can create and distribute these powerful technologies at a high speed and at an incredible scale. But it is the development of a decentralised approach in parallel to this rollout, that will be vital in pushing the technology behind AI forward while ensuring a democratic, and diverse AI landscape.
Why is it important to democratise AI?
For many of the same reasons that democracy is important in human societies, so is it in digital spaces. Authoritarian rule has historically been met with resistance, for failing to ensure equitable development and efficient growth. A monopolised AI system (much like an authoritarian government) may run into similar evils over time - stagnation, inefficiency, and an inability to freely pivot and develop as necessary.
These are, of course, concerns relating primarily to the progress of the technology itself. The more pressing concern, however, has to do with security. Who will come to define AI’s parameters is a vital question - will it be a single entity, or a collective of contributors? A single corporation or a syndicate of Big Tech dictating the security protocols of a worldwide technology has evident, and severe drawbacks. Consolidating that much power, which transcends international borders, and can reach individuals at levels never-before-seen, into the hands of a single firm, could lead to (and has, historically led to) manipulation, inequitable treatment, and even potential persecution.
That’s where the decentralised development of AI comes in - it ensures a more beneficial experience for users from different demographics and regions. By making AI technologies accessible and adaptable, we can promote equitable growth and development - a decentralised system evidently is more able to adapt and evolve to the preferences or requirements of users from different regions or cultures, as it has the accommodation of multiple perspectives built-in.
Community-driven development is vital for the AI ecosystem - and for the democratisation of AI itself. This diversity prevents extreme situations and promotes a more resilient and adaptable AI system. A decentralised resource management, for example, where computing power is derived and managed by the community, will be less reliant on the operations of a centralised system, affording it the resilience that cannot be achieved otherwise - as we’ve found (and continue to find!) at NetMind. At the agent level as well, leveraging a variety of models, each with their core language and perspective built into them, will better reflect the diversity of those making up the real world. The AI community in this sense mitigates the risks associated with any single entity holding unrestrained power over AI.
As AI technologies become more integrated into everyday life, general AI literacy will naturally improve. Similar to how smartphones consolidate multiple narrow tools into a general-purpose device, AI will evolve to serve broader purposes. This integration will reduce the need for narrow AI tools, promoting a more holistic understanding of AI capabilities. As AI evolves, we’ll see a generation that is inherently prepared to engage with and develop these technologies, and interface with them more regularly in daily life. As this reach grows, too, and AI usage becomes increasingly commonplace, it will become more important that there are meaningful checks in place to create an inclusive and robust AI system.
Where next in the AI journey?
The journey towards AGI is marked by continuous advancements, challenges, and opportunities. By democratising AI, we can ensure that its benefits are distributed equitably and that its development is guided by a diverse collective. As AI technologies continue to evolve, fostering general AI literacy and promoting inclusive participation will be crucial for ensuring a resilient and adaptable future. The vision of AGI is not just about technological progress but about creating a democratised, decentralised AI landscape that works for all.
Henry Huang is CTO of NetMind.AI.
