Where is India in the fourth Industrial Revolution — artificial intelligence, 5G, cyber-espionage and drone warfare?
There has been a good deal of talk lately about the fourth Industrial Revolution. Apparently a term popularised by Klaus Schwab of the World Economic Forum, this meme, while intended as catchy marketing jargon, has a certain element of truth to it, that deserves further study. In particular, the negatives implied may well be worth deeper consideration than it appears it has received in India.
To start with, a general definition. The first Industrial Revolution replaced animal and human muscle power with that of machines, eg. the steam engine, the steam locomotive, and mechanised looms. It enabled a quantum leap in the scale at which things could be built or run, as mechanisation overcame inherent limitations of biological entities.
The second Industrial Revolution was the advent of electricity and mass-production techniques. Specialised factories running on the virtually unlimited power of electricity could scale up again by a quantum leap, and thus products could be brought out in the millions of units, thereby making many things that were once luxury items newly affordable and indeed commonplace.
The third Industrial Revolution was that of the advent of computing: first, mainframes, then PCs, and then the internet. Staggering advances in the power of computers and their ubiquity meant that all sorts of things could be automated and run more efficiently, and unimaginable connectivity, entertainment and white-collar job enhancement have been possible, as well as the creation of rich virtual worlds.
Taking the analogy one step further, the fourth Industrial Revolution centres on three things: a tsunami of data, the arrival of deep (albeit narrow) intelligence in machines, and the dawning capability of using both these to influence the physical world, as sensors and distributed devices become ubiquitous.
Each of these revolutions has affected society as well as the planet in innumerable ways, some good, some bad: for instance, pollution and deforestation; the creation of industrial ghettos and mind-numbingly boring jobs; the breakup of joint families and the creation of alienated individuals with the ability to inflict violence on a much larger scale than earlier.
But nations have also suffered by missing the boat, as it were, in these revolutions. Those familiar with the work of economic historian Angus Maddison are aware of how, roughly coinciding with the first Industrial Revolution, the concentration of wealth in the world moved sharply away from the Indo-Pacific (broadly India and China) to the Atlantic (Europe and America). This brings up a broad question about what is happening today.
The answer, to put it bluntly, is that India is missing the boat. Again. While China is not, and is in fact pretty much on the leading edge. This is a dramatic turnaround from a situation in which both India and China were being exploited by the West, to one where India will be exploited by both the West and China, in a worst-case situation. Have things reached a tipping point, and is India doomed? Probably not, if we take steps on a war footing, instead of fixating on silly things like, to take a random example, Supreme Court judges bickering (they should fix it in camera).
As a techno-sceptic, I tend to focus on the negative outcomes of technology, but I have to admit that a lot of good things have come from technology that I too would be loath to do without: for instance, I am writing this on my laptop while on a flight to Paris to attend this year’s UNESCO/Netexplo conference on digital innovation. While yes, I am in a cramped seat with little room for my (small) computer, the fact that I can do this is (and even get access to the internet while on board) is rather amazing.
This means there are lots of opportunities for entrepreneurs to meet the spoken and unspoken needs of consumers, and also for governments to bring in wise policies (because pure market competition is a fiction, as David Teece of Berkeley argued in a seminal paper — industrial policy needs to guide the market with a long-term perspective about core competencies).
But overall, I am concerned about the big picture. With the small advantage gained by their early access to technology (they invented many of the early machines of 1750s), Europeans, and their “guns, germs and steel” (in Jared Diamond’s evocative phrase) were able to enslave us and steal $10 trillion. The consequences would be even more horrific if the Chinese (who swallowed and devastated the massive buffer state of Tibet and are now right on our borders) were to dominate us using digital technology and machine intelligence. And they show every sign of being able to do so, if we don’t wise up, and we need to do it today.
First, data. Indians may have some competence in software, but the very idea of software has changed. We did well to build software “factories” like the British built power looms (textile “factories”) in the 1700s and 1800s. But today to be good at software, we need to be able to understand data, and that too in deep ways, and we are not there.
We do generate data. That is not the problem. With bank automation, digital transformation, e-commerce, smartphone adoption, and Aadhaar, India is generating plenty of data. Given India’s massive scale, I suspect only China and the US are generating more than we are. Unfortunately, we are also giving this data away to others, free of charge. One could argue that our tradition has been to invent amazing things and put them in the public domain, which is true, but given the massive fortunes being made by keeping data proprietary (just see how Tencent and Alibaba have now become global firms, primarily because the Chinese government keeps Google and Facebook and Amazon out of China), this is a bad tactic.
What is the solution? I don’t know if India should go so far (and indeed I do not know we actually could) as to force Google, Facebook, Amazon, Tencent, Alibaba et al out of the country. But one possibility is to insist that all of them maintain their Indian data on Indian servers. If I am not mistaken, the European Union insists that European data must reside on European servers, and not be diverted to American servers. Purists will scream that this damages the open nature of the internet, and it does; but so does the proprietary walled garden Facebook has created. That would be a start, instead of Alibaba using subsidiary Paytm’s data as it pleases. I also don’t think many Indians would want our medical data or financial data being owned by foreigners (surprise: credit agency CIBIL is owned by America’s TransUnion).
The second moonshot area is artificial intelligence (AI). I read about how NITI Aayog has been funded with Rs 3,700 crore to create a focused effort in this area, which is a start. But look at the scale of what China has already done. I am quoting from “China seeks dominance of global AI industry”, in Financial Times, 15 October 2017. China has a plan to create a $150 billion AI industry by 2030, although Kai-Fu Lee, formerly head of Microsoft Research and Google in China, and now a venture capitalist in AI, says it will happen sooner, given the reams of data, talent and investment.
The scale of Chinese ambition is high. Says a government policy paper, “(China) must, looking at the world, take the development of AI to the national strategic level... firmly seize the strategic initiative in the new stage of international competition in AI development, to create new competitive advantage, opening up the development of new space, and effectively protecting national security”. Do note that last bit: it’s not only commerce, they have warfare too in mind.
AI, according to the boosters, is going to be one of those transformational technologies that will be in absolutely everything. Here is where India’s lack of R&D is really hurting us. Not having a steady stream of high-quality PhDs and papers coming out of labs in universities and in industry is a big problem. One short-term solution would be to set up truly autonomous AI labs and lure away Indian-origin researchers from the West to do their work, unhindered by the usual busybody Indian bureaucracy. (China did this, and, as a result, have now overtaken the EU in the number of papers published on AI). Long-term, we simply have to improve the universities.
The third kingpin is communication. With the expected boom in internet-of-things devices in everything from medical devices to pollution monitors to self-driving cars to smart sensors in fields, there needs to be a “fat pipe” to vacuum up all the data to the cloud and also to deliver content to the ubiquitous devices. 5G telecom, with ample bandwidth, is the answer that most are proposing. This is so important that there was a stir recently when a leaked white paper suggested that the US should consider building, at government expense, a 5G backbone, instead of what it normally does: allow various private players to do it themselves.
There is a high-level task force in India that is looking at 5G rollout, but I wonder if it sees the big picture strategically: that it is, among other things, infrastructure for national security and for cyber-warfare.
That brings me to the final aspect of this revolution that makes me wake up in a cold sweat: cyber-security and the use of robots and drones for warfare. The civilian infrastructure that is being put in place by China in particular is easily adapted to warfare. For instance, China is the leading manufacturer of small drones in the world. But by adding machine intelligence and small arms to these drones, China can change the very nature of warfare (even if you leave cyber-attacks aside).
It can be a game-changer, as guerilla warfare has been, or German blitzkriegs. Or see how the English invention of the longbow decimated French cavalry charges.
Imagine a swarm of small drones (<$1000, off the shelf) in an ad hoc mesh network (that is, they communicate with each other and coordinate in a loose federation) attacking an Indian position, in, say, Doklam. This is not even theoretical. Home-made drones have been used to devastating effect by Islamic State in Syria and Iraq. If they carry something as simple as a grenade, they can be deadly: I read somewhere that the US military manual suggests withdrawal as the only solution, as these things are hard to shoot down with rifles, and they are too small to show up on the imaging software for bigger guns.
Now imagine a thousand of them coming at you, and with AI, they have the emergent intelligence of crowds (imagine a beehive). They would be impossible to defend against, like a swarm of locusts. Or a swift Mongol horde on horseback: a swarming tactic was what Genghis Khan used to conquer half the world. Although he was a Mongol, China now claims him as theirs, and it shows. They are already using swarm techniques at sea (“fishing boats”) and on land (“shepherds”) to create facts on the ground.
What is our innovation against this? Similarly, what is our innovation against cyber-warfare and asymmetric warfare, which Chinese military planners have spoken of in detail? What are all those Chinese-designed smartphones doing? Can someone reverse-engineer their code and figure out if all your most confidential conversations are ending up in some server in China? If China wanted to wage a small and undeclared war, what would Huawei’s switching equipment and all the gas turbines and other power plant equipment do? Does the Chinese army have the kill-switch?
India had an industrial policy that was meant for state control of the “commanding heights of the economy”. As expected, it was a disaster. But a more flexible, nuanced type of industrial policy — perhaps like the US does with DARPA (Defence Advanced Research Projects Agency), and Japan did with MITI (Ministry of International Trade and Industry), and what the Chinese do with their focus on specific strategic technologies, would not be a bad idea for India now. And it fits in squarely with C K Prahalad’s superlative idea of “strategic intent”.