Digital Native: Cause an Effect

The article was published in the Indian Express on June 17, 2018.

Statistically, it has been proven, that the consumption of ice cream in the country increases significantly in the summer months. In the same months, the number of housebreak incidents also increase. It might be possible, though ridiculous, to now make an argument that eating ice cream leads to increased frequencies of housebreakings, and, hence, sale and consumption of ice cream should be regulated more rigorously. The humour in this situation arises out of the fact that we know, at a very human level, that correlation is not the same as causation.

We know that just because two things happen in temporal or spatial proximity with each other doesn’t necessarily mean they are connected or responsible in a chain of events. This is because human communication is designed to make a distinction between cause-and-effect relationship and happened-together relationship between two sets of information.

However, when it comes to computation, things turn slightly different. Within the database logics of computation, two sets of data, occurring in the same instance, are subjected to a simple scrutiny: Either one of them is linked with the other, or, one of the two is noise, and, hence, needs to be removed from the system. Computation systems are foundationally anchored on logic. Within logical systems, all the events and elements described in the system are interlinked and have a causal relationship with each other. Computational learning systems, thus, do not have the capacity to make a distinction between causal and correlative phenomena.

This is why computation systems of data mining and profiling are so much more efficient than human cognition. Not only are these systems able to compute a huge range of data, but they are also able to make unprecedented, unforeseen, unexpected, and often unimagined connections between seemingly disparate and separate information streams. I present to you this simplified notion of computer logic because it is at the heart of the biometric identity-based debates around Aadhaar right now. Recently, Ajay Bhushan Pandey, CEO, UIDAI, wrote an opinion piece that insisted that the data collective mechanisms of Aadhaar are not only safe but also benign. His opinion is backed by Bill Gates, who also famously suggested that “Aadhaar in itself” is not dangerous.

And, in many ways, Gates is right, even if Pandey’s willful mischaracterisation of Gates’s statement is not. For Gates, a computer scientist looking at the closed architecture of the Aadhaar system, it might appear, that in as much as any digital system could be safe, Aadhaar is indeed safe. In essence, Gates’s description was, that as a logical system of computational architecture, Aadhaar is safe, and the data within it, in their correlation with each other, does not form any sinister networks that we need to worry about.

However, Pandey takes this “safe in itself” argument to extend it to the applications and implementations of Aadhaar. He argues that because Aadhaar is a self-contained safe system, its interaction with other data and information systems is also equally safe and benign. In this, Pandey, either out of ignorance or willful mischaracterisation, confuses correlation with causality. He refuses to admit that Aadhaar and the biometrics within that are the central focal point around which a variety of data transactions happen which produce causal links between disconnected subjects.

Thus, the presence of a digital biometric data set might not in itself be a problem, but when it became the central verification system that connects your cellphone with your geolocation data, your presence and movement with your bank account and your income tax returns, your food and lifestyle consumption with your medical records, it starts a causal link between information which was hitherto unconnected, and, hence, considered trivial.

The alarm that the critics of Aadhaar have been raising is not about whether the data on Aadhaar is safe or not, but, how, in the hands of unregulated authorities, the correlations that Aadhaar generates and translates into causal profiles have dire consequences on the privacy and liberty of the individuals who carry the trace of Aadhaar in all facets of life. Pandey and his team of governors need to explain not the safety of Aadhaar but what happens when the verification information of Aadhaar is exploited to create non-human correlations of human lives, informing policy, penalisation and pathologisation through these processes.