As we move further into the Fourth Industrial Revolution, the role of data becomes increasingly important. Just as data is exploding in volume from the internet of things, mobile and other sources – so-called Big Data – so too is the pace of the technical transformation expanding at an exponential rate, making our interaction with machines and information more common, natural and powerful. Data is the basis for many revolutionary AI applications, from gene-sequencing to robotics, to modelling climate change, developing autonomous vehicles and improving agricultural yields.
Subsets of AI, machine-learning algorithms and especially deep learning neural networks require huge datasets for training. The more data, the more accurate their predictions. The more X-rays of tumours a neural network can analyse, the more likely it will accurately characterize the next one it is shown. Less data means less accuracy – which, realistically, means a disadvantage.
Have you read?
Today a few large corporations are favoured, as they are the best positioned to collect and process vast quantities of data from ecommerce, digital assistants and other sources. Several countries also have inherent advantages due to their sizeable populations, with vast pools of collectable data about many facets of life, from driving behaviours to cell phone use to internet browsing. This leads to an imbalance of power and wealth, caused by information being in the hands of the few, which gives them the opportunity to use these large volumes of data to draw meaningful inferences and achieve economies of scale.
What about countries with smaller populations, or without the resources to develop state-of-the-art technologies? Some of these countries may be small but are otherwise wealthy; they’re only disadvantaged by a lack of data. Other countries may be historically behind on common measures of development. All of them are at risk of falling from their current positions simply by lacking accessible and relevant data. Countries without this data have the risk of being left out. They would be relegated to providing the raw resources of production without any actual means of production.
As with trade agreements for goods and services, one way to help less advantaged countries and companies is to create data-sharing agreements or pacts. As with trade for goods and services, there are advantages for all. Combining data from multiple countries has the potential to help each of the individual countries achieve the required critical mass of information to fuel advanced applications. In some instances, the agreements could be worldwide. Cancer, for example, knows no borders. In other instances, if countries share a common language or culture, a regional approach could be more appropriate. Already, PwC expects the global data economy to be worth more than $400 billion by 2025. Access to more data could substantially expand this by encouraging more startups, more investment and lead to additional breakthroughs to improve the quality of life across the planet.
Putting a price on our digital dust
Expanded access to data could be ideal for all countries. However, the data originates with individual people. Their data individually and collectively is being used and will in the future be used in many instances for commercial purposes, and consequently has value. People should be able to share in the wealth that is created from their data. They should be compensated for this use.
Privacy laws today vary from country to country, and make the collection and dissemination of data across borders very difficult. Guarding privacy is clearly positive. However, if an individual chooses to provide their genetic data and health history for medical research, a mechanism to do so needs to be created for their benefit and for the benefit of larger populations. Of course, the purpose doesn’t always have to be as grand as medical research. The information could simply be our digital dust – including social media posts, the fitness analytics from our smart watch and the driving habits of everyday life.
One possibility to realize this vision is to create an exchange to determine the value of data, much like commodity exchanges. If someone – a country, an alliance of countries, a lab, an enterprise – wants to use data, they could offer tokens as a payment proxy. An exchange could establish the price of the tokens. As data exchanges publish value, people could see the value of their data and make an informed decision about whether to provide this to the exchange. This also needs to be complemented with “intended purpose”, as opposed to the historical trend of “source of origin” regulation of data, which may be enforceable through, for example, blockchain smart contracts.
Though still at an early stage, pilots for this form of data-sharing and value exchange are underway. If they show promise, then an entire industry of data-sharing could be created. While this has the potential to help everyone – just as with trade agreements – the benefits could be the greatest for developing countries: what I call “AI for AID”. By leveling the AI playing field through these data exchanges, the benefits of the Fourth Industrial Revolution will spread more evenly across countries and businesses, creating new markets and driving economic growth.