Why we need an Open Data Economy

Written by Laurent Rochat, Managing Director of Innovation Atelier Ltd and Ocean Protocol Ambassador. This article was originally published on LinkedIn.

Over a decade ago, when I was a student, and when statistics revolved around linear regressions in all their straightness, I remember very clearly of our professor highlighting how computer processors and statistical packages were so powerful now, to a point it was getting dangerous because anybody was now able to run advanced statistics without understanding any of it.

Today, everybody uses them, stats are everywhere. Most people still don’t understand any of it, but progress has been tremendous. Things that were barely conceivable at the time, like cool chat bots and safe self-driving cars, have become a base expectation for everybody. These AI-based solutions are getting implemented in every aspect of our life, from answering voice calls to opening doors.

AI-based solutions are getting implemented in every aspect of our life

And today, anybody can build an AI model using open-source deep learning libraries and see the results for themselves. Hardware and AI libraries can compute models of much higher complexity about 10’000 times faster than old school statistical methods back in the days. And the output of these models is not just informational, it is experiential.

Today, everybody can use the output of these models as a component of real-world user applications. Ironically enough, nobody, even the most brilliant AI scientists, can understand exactly how these models work.

Thinking about this puts a smile on my face. We were so naïve at the time. Right?

Well, in fact I’m pretty sure that I will smile again in a few years as I read this essay again. We were so naïve in 2019. Because technology and society will keep evolving at a steady pace, and our ability to apprehend such evolution will keep lagging behind.

Deep learning has this particularity to only get better when more data is available

Now, improvements in computation capabilities and algorithms are only partially explaining the boom in ML and AI we have witnessed in the last few years. AI, and particularly the most advanced stuff like deep learning, has this particularity to only get better when more data is available. And data production has boomed in the last decade. Consider that in 2013 already, 90% of all the data available was generated in the last two years, a time when the internet of things was not even a thing yet.

In 2013 already, 90% of all the data available was generated in the last two years

This is mind blowing. To the extent that experts view data as the new oil. And indeed, data has been fueling most of the economic growth measured in developed economies. But unlike oil, data is not consumed. It is kept in well-guarded silos. It has too much value to be shared, because sharing data is viewed as a one-way process, which may forever jeopardize the competitive advantage (and the intrinsic value) of single-handedly owning that data.

This is problematic, and the problem is multi-faceted.

• On the one hand, data provides too much power to the few people who have both the data and the know-how to use it. This now represents a significant source of concern for the UN.
• On the other hand, much data just sits there with no productive use. Many non-tech companies of all sizes sit on relatively big amounts of data, which they do not currently use to their full potential because either they don’t have the skills to use the data, or because the data is too small, too specialized or too narrow to be exploitable.

The big opportunity lies in breaking silos of this data that is not used today

It would be politically risky to impose anything to big tech behemoths, so they are likely to keep expanding and enriching their data as they see fit. The big opportunity lies in breaking silos of this data that is not used today among all other players.

Until recently, we did not yet have the tools to seize this opportunity. But the day has come thanks to the emergence of two other phenomenons that have shaped society over the last decade: the uberization of work and the emergence of distributed ledger technology.

• The success of Uber tells us something about the power of markets; no one was able to compete with a market of loosely coordinated independent players. Adam Smith had predicted that an invisible hand was driving markets in the 18th century, and it has pushed Uber to the success we know today. Uber business practices are debatable, but the success of their model is undeniable.
• The rise and longevity of Bitcoin shows that money is a social construct; as long as one knows that someone else will accept something as a mean of payment, one can safely accept that something oneself as mean of payment, even 32-bytes strings. It started with a handful of people in 2009 and grew exponentially from there thanks to network effects. Bitcoin has also shown that distributed ledger technologies (DLT) can create value among counterparties that do not trust each other. This enables collaboration between counterparties that would otherwise only be able to compete with one another.

Distributed ledger technologies can create value among counterparties that do not trust each other

The combination of these two phenomenons has given birth to a new paradigm: one can create markets among competitors in the same sector, no matter how fierce the competition. In fact, the oil industry is already using blockchain technology to manage collaboration on pipelines usage. They may compete on all aspects of the business, but when it comes to logistics it’s much more effective to use a clear set of rules (and a can’t be evil moto) than a business association to manage dispute (in a shouldn’t be evil modus operandi). It works in the oil industry, but also in and many other industries where DLT is changing collaboration in logistics/shipments, accounting, finance, marketing …etc.

Of course, this exact same model can be applied to data and analytic services. And this is precisely what some smart people are doing: Ocean Protocol is building a new open data economy. They are creating the tools to be able to match data providers with data scientists. They are creating a reference market for curating and exchanging data. They are creating the future of data economy where data and skills can be monetized independently in joint, yet trustless value creation.

Ocean Protocol is building a new open data economy

Participants in the market are incentivize to share the data they have with others. Concerns over disseminating ownership of the data is addressed by integrating on-premise computation, meaning that no data ever has to leave the servers where it is located, mitigating the risk for data owners. The specifications of these markets are still being carved out to date, but it makes no doubt that when the market is ready, it is going to have a huge impact on the way we create and share value from data.

Freely flowing data across teams and structures will unlock new insights. It will even up the game between tech giants and all others. It will enable new services and new business models. It will change the way 99% businesses operate today.

Currently a consortium of car manufacturers is pooling their driving data using the platform to be able to train a self-driving car model, in catch-up mode vs Google, Apple and Uber. But not for long in my opinion. I firmly believe that when we will look back in 10 years, we will observe that those who did not piggy back on the power of this nascent data market, no matter how much ahead they were at the start, have been left behind. The invisible hand of the market will have worked its magic, again.

This is my hope, at least. But I don’t want to be hopeful, I want to be part of it.

More information can be found at https://oceanprotocol.com/

Interested? Join us at https://www.meetup.com/Data-Economy-Geneva/

Ready to dive in? — Check out the Ocean Documentation to get a sense of what we’ve been working on, and fire all your technical questions in our Gitter chatroom.

Follow Ocean Protocol on Twitter, Telegram, LinkedIn, Reddit, GitHub & Newsletter for project updates and announcements.