Imagine a world where you could share data freely, and everyone would use the data appropriately. What would you be able to do? Much as the tragedy of the commons is a classical story of a failure of entities to self-regulate, groups struggle to share sensitive data for fear of their interests being violated. There are seemingly insurmountable obstacles in data sharing today. But, what would you be able to do if the risks were accounted for?
We dare you to ask this question and imagine what your entity might become. We know that today data is the most valuable resource on the planet; more than any natural resource. Yet, similar to petroleum resources, the richest reserves are difficult to get to where they can be used. Now, this is where the analogy breaks. For as long as we have had data the operating model has always been to bring data to the analysis. When you flip this on its head you start to find inherent protections for data holders. While California Privacy Law addresses movement and exposure of data, there is no representation for what is needed if you bring machine learning to the data and run your algorithm double-blind. If the data owner can't see the algorithm and the algorithm owner can't see the data, we step beyond what regulators have been able to envision.
Today, algorithms need raw data to draw insights, yet there is no need to expose raw data to humans. Now, how much faster can you create data-sharing agreements when data is neither exposed nor moved? What new uses become available? What happens when you are able to leverage data that was previously too sensitive or risky to properly utilize?