Anyone who has a substantial reservoir of data can reasonably expect to derive powerful insights from it. These information are often not used to increase advertising revenue or ensure greater customer stickiness. In other instances, our political preferences have been distorted to change and manipulate us into making decisions that we might not have otherwise.
The ability to generate insights keeps the people who have access to these data sets, whose data is contained within them. This allows the former to benefit from the data in a way that the latter may not have thought possible even if the latter had consented to provide it. Given that these insights can be used to harm the people to whom it relates, the effects of this data asymmetry need to be mitigated.
Privacy law attempts to do this by providing data principals with tools they can use to exercise control over their personal data. It requires data collectors to obtain informed consent from data principals before collecting their data and forbids them from using it for any other purpose than has already been notified. That is why, even if that consent has been obtained, data trustees cannot collect more data than is absolutely necessary to achieve the stated purpose and are only allowed to retain that data for as long as that is necessary to fulfill the stated objective.
In India, we have gone a step ahead and created techno-legal solutions to help reduce this data asymmetry. The Data Empowerment and Protection Architecture (DEPA) framework makes it possible to extract data from the silos in which they reside and transfer it to other entities on the instructions of the data principal, who may use it to provide other services to the data principal. . , This data micro-portability undermines the historical advantage that executives accrue by collecting data over the entire span of their customer engagement. It eliminates data asymmetries by establishing a competitive market-making infrastructure for data-based services, allowing data principals to choose from a range of options as to whether their data will be used by service providers to their advantage. how can be done.
However, this is not the only kind of asymmetry we have to deal with in this age of big data. In a recent article, GoLab’s Stephan Verhulst at New York University pointed out that it’s no longer enough to have large repositories of data—you have to know how to effectively extract value from it. Many businesses may have vast reserves of data that they have accumulated over years of operation, but very few of them are able to effectively extract useful signals from that noisy data.
Only owning a large data set is of little value, without the knowledge of how to translate the data into actionable information.
Unlike data asymmetries, which can be reduced by making data more widely available, information asymmetries can only be addressed by fundamentally democratizing the techniques and know-how needed to extract value from the data. This information is largely proprietary and difficult to access even in a perfectly competitive market. What’s more, in many instances, the computation power required exceeds the capacity of entities for which data analysis is not the main objective of their business.
That said, we’ve started to see new marketplaces in recent times that address this need—platforms that provide access to off-the-shelf models and AI algorithms that can be used to generate data from a range of different data sets. can be done to extract the value. The resulting democratization of data science has made it possible for ordinary businesses to extract value from their own data that was not possible before. This, in turn, has begun to address the information asymmetry that separates people with technical knowledge from those with data.
Before closing, there is another type of asymmetry that is often discussed in the context of data. As technology improves, decisions made on or based on suggestions from algorithms will affect us in increasingly important ways. Today, AI is used to determine our eligibility for loans, the value of our insurance premiums, and even the nature and duration of prison sentences. Each of these decisions has an impact on the lives and livelihoods of ordinary people – and any bias inherent in the algorithm could unfairly prejudice the people to whom the decision applies. This is especially true for so-called ‘black box’ algorithms in which the rational for the decision remains opaque even to the operator of the program. This inability to understand how automatic decisions are made is what Verhulst refers to as intelligence asymmetry – and this needs to be addressed to prevent harm due to algorithmic bias.
Regulatory response to this has traditionally required that automated decisions be accompanied by an explanation of the basis on which that decision was made. But interpretability is often a trade-off against accuracy. Algorithms whose decisions can be interpreted are often no less accurate than those whose decisions are inexplicably—in a black-box.
But is this a compromise we are willing to make? What if a black box algorithm could accurately diagnose a lesion on your skin, sooner than any human radiologist could hope. Would you still insist that life-saving algorithms should not exist because their decisions cannot be explained?
Rahul Mathan is a participant in Trilegal and also a podcast called Ex Machina. His twitter handle @matthan . Is
catch all business News, market news, today’s fresh news events and breaking news Updates on Live Mint. download mint news app To get daily market updates.