Fetch.AI — A new framework for unlocking the value of data

According the Huawei’s Digital Spillover report published in 2016, the true size of the data first, digital world is about $12.9 trillion or roughly ~18% of global GDP.

Unlike the natural resources like oil that drove previous paradigm shifts in the economy, the new resource of data is not truly scarce in nature or geographically limited. This forces us to reconsider how businesses collect, retain, share, trade and analyse data. If the intellectual property rights of the 20th century hint towards anything — it is that “intelligence” in the previous economy was limited in access to reduce counterfeits.

However the rise of open-source communities in the web and data share agreements between enterprises (eg: Facebook & Huawei), point towards an economy where returns compound as a result of data being actively shared between players. Present day solutions do not allow that to occur at scale.

Data collection is broken, its storage is centralised, its sharing is monopolised and its trade has no requisite infrastructure in place. It is almost as though, we have built the trains that will deliver economic value in the 21st century, but forgot to build the rails for them to pass upon. Cheaper computing (credits to Moore’s law) combined with mankind’s tendency to aspire for efficiency through automation has resulted in the creation of a resource that is not optimally used — Data.

Fetch aims to solve this problem by creating a new model that involves autonomous agents working anonymously on behalf of the individual without leakage of personal data.

In simpler terms — Fetch would create digital agents that work anonymously on behalf of an individual in a virtual economy to find, interact and trade upon data sets that may be beneficial for an average consumer and all while putting them in complete control for the data they own.

Data being stored in silos by a handful of enterprises reduces its utility. The source of data, be it an individual’s spending habits or sensors owned by a business, can be empowered through digital agents that are intelligent enough to advertise and trade it for value that is useful. Much like efficiencies in engines and connectivity played a role in enabling the industrial revolution, improvements in how data interacts with each other could be detrimental to unlocking value that currently remains locked.

Even in the presence of data exchanges, the cost of data curation and advertising may often be more than the value created by trading it. This is why intelligence is needed in these contexts. A digital first agent can interact with thousands of other agents, each with unique data offerings to find what may be most relevant to the user.

Fetch would create a parallel, digital-first representation of the world in which autonomous agents can interact with each other to do useful work. In this case, “useful” work could mean finding information a business could use to optimise its functions, offers or deals an individual could avail or variations in sensor data from a remote business on the other side of the world. All without the involvement of a human being.

This would occur within what the we call an “Open Economic Framework” (which we also refer to as the Fetch ‘Digital World’). This is a market that is devoid of both geographic boundaries and national borders that currently limit value transfer; one that enables discovery and trade of data that would otherwise lie unused.

In Fetch’s case, the three foundational work-pieces are optimising network performance, delivering trust information and creating, refining and using market predictions and correlations. It is this last one, predictions, that is so fundamental in making the Fetch world a place where genuinely useful, profitable value exchanges can take place with such ease and lack of friction.

These predictions not only allow real-time restructuring of the world to clear away the junk and focus attention on what is, or might be, important but they give agents access to market intelligence that was previously locked up tightly in centralised data silos owned by the few and unavailable to the many.

Present day economies are largely designed to be optimised for human beings. We network, find and create value together in human intelligence. Even in the presence of digital platforms like LinkedIn, we rely upon innate abilities to collaborate. It is indeed this ability to work together that took us from hunter-foragers to a species capable of landing on the moon.

As society evolves to be machine driven, the requisite for creating an ecosystem that allows intelligent agents to do the same increases more than ever before. This is at the crux of what Fetch aims to solve through unlocking data that may seem useless at first glance and allowing agents to interact with each other for its trade and utility.

Unlike humans, data has no limit in terms of migration or speed of transfer. It can be literally transferred at the speed of light due to fibre optics. However, they lack the means to structure, explore and find value within themselves. Given the amount of power spent on mining on various blockchains in this day and age, Fetch provides an alternative where the power expenditure can be used towards machine learning and in extension, what we call useful proof of work (uPoW).

The ledger repurposes the processing power required to reach consensus for doing genuinely useful work: and that shifts a human-driven economy into one that’s digital, a journey from a mining ecosystem that creates little value to one that can be theoretically the most powerful machine learning system in the world.

That’s what Fetch aims to achieve over time: collective, constantly evolving intelligence for the benefit of all its users. We will be exploring each of the components that allow Fetch to achieve this — in an oncoming series. Stay tuned!