Looking Back on 2023

2023 was the year in which DPI assumed its rightful place on the world stage. It was also the year in which artificial intelligence came into its own. There has never been a more interesting time to be engaged in technology policy.

This article was first published in The Mint. You can read the original at this link.

This is the last Ex Machina article of the year and, as in years past, I wanted to use the opportunity to reflect on the year that was.

Interest in DPI

Last year at this time, I had begun to notice a buzz developing around India’s digital public infrastructure (DPI) - not just within the government which, despite being at the heart of many of these developments, seemed unaware of the sheer magnitude of what it had achieved - but also more broadly around the world that had slowly started to appreciate exactly what it was that India had built. In my year-end article, I hinted at how India’s Presidency of the G20 might be a timely opportunity to showcase the unique techno-legal approach at the heart of our DPI approach. Given the building momentum, I was cautiously optimistic that the idea might catch.

If anything, the events of the past year have far exceeded my expectations. Not only did India make DPI one of the central pillars of its Presidency, it built such broad consensus around the approach, that it was unanimously endorsed in the significant outcome documents of the Indian G20. What’s more, through bilateral and multilateral engagements over the course of the year, various countries have committed to work with India to build DPI, both in their own countries as well as to jointly deploy them in developing nations that need them. It would not be an exaggeration to say that this was the year that DPI finally assumed its rightful place on the world stage.


We now face something of a different challenge. Having convinced them of the many benefits DPI can bring, countries big and small are lining up to seek our help to deploy DPI for them. It is becoming evident that as powerful as the idea is, deploying DPI in another country is not a simple matter of simply copying the protocols and the code that worked in India onto a server in another country and switching it on. For DPI to work it needs to address the unique demands and constraints of each jurisdiction in which it is being deployed.

Much of the coming year will be spent figuring out how to most efficiently globalise DPI. We will need to learn how to package the core building blocks of the DPI approach into cloud-based offerings that are each rapidly deployable - so that we can, pretty much on demand, spin up an instance and do so at an affordable price. Once this is possible we will need to develop the skills to repackage these building blocks into DPI solutions that are tailor-made to meet country requirements appropriately customised to account for local constraints. It is only then that the idea of DPI will be globally scalable.

Artificial Intelligence

As much as last year was the year of DPI it was also the year of artificial intelligence (AI). Over the course of just 12 months, the technology grew from a clunky, wildly inconsistent dalliance to a virtually indispensable piece of technology capable of being used across a number of different use cases. Small wonder that for every DPI article I wrote last year, I wrote 2 on AI.

Today, AI is integral to my workflow. I use it for research - to understand difficult concepts and annotate and summarise lengthy texts so that I can cross-reference ideas and concepts across my notes. I use it to build outlines for my writing (including articles like this one) and, once finished, to generate AI art to illustrate them. I’ve used it to extend myself beyond my comfort zone, taking its help to hand-code my own static website and build small software applications even though I have no knowledge of the underlying programming language.

But despite the many benefits AI offers, much of the conversation around AI this year has been about the societal disruption it is starting to cause. We have long struggled to come to grips with the bias implicit in these models, the fact that copyright and notions of intellectual property are utterly meaningless to these technologies and the harms that they will cause due to the unconstrained proliferation of deepfake augmented disinformation. Jobs in every industry are at risk as employers look to drive efficiencies at all levels by using AI wherever possible and parents everywhere are scared about how this will affect their children - if an AI chatbot can be made to write school essays, how will they ever learn?

Societal Backlash

Last year we saw the beginnings of a societal backlash against AI. Actors and screenwriters in Hollywood went on strike against the increased use of AI in the film industry. Employees in other industries are bound to follow suit. Schools and colleges cracked down on students who were using AI for their submissions, using AI detection technologies to identify whether the assignments had been written by a machine or not and eventually courts and other institutions will as well.

The issues thrown up by both DPI and AI will play out over the course of the coming year. We will start to see vocal pushback from those in society who see both forms of technology disruption as a threat to their existing way of life - just as we will see those who benefit from all that these technologies have to offer urge faster and more aggressive adoption.

Technology policy will have to walk a tightrope between these two extremes, finding a way to maximise the benefits of technology while still minimising the harms. It will call on us to be innovative about the laws we design and empathetic to those whose ways of living will be affected.

There has never been a more exciting time to be in Tech Policy in India.