🕊️Does Data Need Nirvana?
Networks, by Alina Grubnyak on Unsplash
Summary
Data is playing an increasingly seismic role in our economy. It underpins everything from targeted marketing to the development of artificial intelligence.
Using blockchain technology as the foundation for a transparent and secure ‘digital playground’, there is an opportunity to return data – including the innovation it drives and wealth it brings – to the hands of its creators… the public.
The Problem
You may have heard it before. Data is the new oil. Though the wealth is also being concentrated in the hands of the few, unlike oil, data is something we all produce and – in theory – own.
However our emerging data economy does little to reflect this transition in ownership. Instead the race for data has driven several key problems:
1. Walled Gardens of Data
Currently data lies across siloed digital gardens controlled, gate-kept and sold by companies. This means datasets are less comprehensive. With the quality of training data being a critical limiting factor in the development of AI, and publicly available data nearly all used up, developers worry that we're reaching a 'data wall'.
2. Opaque, undemocratic AI innovation
How our data is used to fuel AI innovation currently lies beyond the hands, and often eyes, of the public. As personalised AI emerges, the transparency and governance of innovation is going to become increasingly urgent. Does the public, being both training-data-provider and target customer, not have a right to participate in the evolution of these technologies?
3. Increasing wealth concentration
Controlled and sold by a select few companies, often for sums between $50-200m, yet produced and owned the collective, the current data economy feels almost feudal. Indeed Reddit are one of the latest to reach a big data agreement, selling their users data to Google for over $60m a year. (Check under ‘For Digging Deeper’ to read about the Reddit community’s response.)
The Scape
Imagine in the not-so-distant future, you have made the leap into the new data movement now over 100 million strong.
Though one of the younger members of the community, you have already set up your blockchain wallet and uploaded various datasets of yours – from your reddit usership to your Instagram activity – to your digital avatar.
Still learning the ropes, you have chosen to remain hosted by Vana – the pioneer (and still the largest) of the 'user-owned data' networks. They encourage you to self-host eventually for maximum security but for the time being your data is encrypted by their server and controlled on your personal server.
You are exploring the different groups and apps you might join and contribute your digital avatar (or data) towards. Called DAO's – decentralised autonomous organisations – these groups are leaderless with all the protocols voted upon by the collective. Some of these are building massive datasets to sell to AI developers, some are building foundation AI models themselves using these comprehensive datasets, and some are building apps designed for personalised AI's.
One of the latter, an Dating app called Flirtual has piqued your interest. It requests that you move your dating app history to their secure, decentralised platform. While your data remains encrypted, its contribution to the ecosystem is rewarded with tokens that allow you to vote on upcoming decisions or monetise. With the help of your personalised AI assistant, you begin to explore the world of private online dating.
Vana emerged as the the first in this space, identifying the opportunity and need to provide a platform that transformed the governance of data, and therefore the development of AI, using blockchain technology.
Blockchain essentially acts as a decentralised, distributed ledger that records transactions of information. It does so transparently, whilst maintaining security, thus allowing for user owned and controlled data-sharing.
Short for Nirvana – a play on the idea of liberating data – Vana developed a complex digital architecture. Their Data Liquidity Layer provides the 'on-chain' infrastructure for data contributors, data poolers (DLP's), and the system custodians and validators to interact in a transparent and secure manner. The Data Portability Layer, built on top of the 'on-chain' Liquidity Layer, allows data contributors and developers to interact around building new apps and foundation models.
Ultimately, this complex architecture comes together to create a transparent 'digital playground' free from the hands of Big Tech. A playground where data holders (everyone) can contribute towards, and get rewarded by, projects that they believe in. Where developers can organise in a decentralised manner to experiment with new ideas. Where comprehensive datasets, able to drastically improve AI innovation, are built and owned by the collective. Where these datasets can be used to train foundation models or sold back to larger tech companies.
Though new to the community, you are part of the final wave need to reach the identified critical mass of 100m users. As founder Anna Kazlauskas said early into the venture,
"A user-owned foundation model trained by 100 million people is not like a YC SaaS company that you sell in six months. This is going to take a while!"
Downstream Value Creation
1. Comprehensive datasets that transcend the data 'Walled Gardens' of today.
Each user, transferring their various datasets held by different companies – think Tescos, Tesla, Twitter – and compiling them on the Vana network, comes together to build far more comprehensive datasets that currently used.
2. Transparent data usage, transparent innovation.
By transferring personal data onto the secure, transparent Vana network, how people use their data is visible and self-determined. Democratising data is a critical step towards democratising AI innovation.
3. Aligned interests across data owners/contributors, developers and consumers.
By ensuring that data contributors have a 'stake in game' and are rewarded for their contribution and participation, contribution is encouraged. Similarly the incentive for the developers and consumers is to build governance and reward systems that attract more contributors. And so the feedback builds.
4. New digital economy.
By returning data to its rightful owner, and then rewarding contribution, a new digital economy begins to emerge. One where wealth flows to the many.
5. Control over AI personalisation.
Personalised AI is coming. Better to own and control what form that personalisation takes and what ends it serves, than to not. As Kazlauskas says, “It would be really scary if someone could take the clone of my voice and use it for whatever they want.”
For Digging Deeper…
The inspiration behind this Scape, Vana is the first ‘User-owned Data Network’ already attracting millions of users.
A more in-depth exploration of the architecture underlying their ‘digital playground’.
In response to Reddit’s sale of users data, this DAO on the Vana network is now the biggest DAO in history encouraging the transfer of user’s data into the collective.
