Arecent Mckinsey Global Institute report, “Artificial Intelligence (AI): Implications for China”, should be of interest because of its pertinence for India as much as China. It appears to have been commissioned by the Chinese counterpart of NITI Aayog and is noteworthy for that reason as well.
State of Play
The Report pronounces authoritatively on the advent of AI (which it defines as “machines mimicking cognitive functions typically associated with the human mind”) globally:
“Advances in data collection and aggregation, algorithms, and processing power” have taken the idea (of AI) from “futuristic speculation to present day reality”. Computer systems no longer need to be “programmed to execute rigidly defined tasks”. They can now “adapt to new data”, without need for reprogramming (explicitly), on the strength of “generalised strategy inputs”.
As a result, many “machine learning systems have already been developed for commercial use” with rapidly growing varied applications in “sectors such as finance, health care and manufacturing”. Recent breakthroughs have pushed boundaries even further, enabling AI systems to “learn, discover and apply rules by themselves” and “match or surpass human intelligence”.
Although machines that can perform the full range of cognitive tasks that humans can do are said to be “still decades away”, the advances made are predicted to have a disruptive impact on all economic activity (world-wide) because of the “dramatic… boost (in) productivity” they are going to bring about in their wake: “AI systems improve productivity by completing existing job activities more efficiently” — they can “make industrial machinery, supply chains, logistics routing and other (management) processes more efficient … by predicting failures, identifying bottlenecks and automating processes and decisions”.
China
The Report estimates that “half of all work activities in China could be automated, making it the nation with the world’s largest automation potential”. It terms that country as “one of the leading global hubs of AI development”, “with its biggest tech companies driving significant investments in R&D”.
But then it goes on to temper that assessment with the observation that “China will need to focus on building its innovation capacity” so that actual adoption of AI manufacturing or other applications is “not just among its technological giants but across its traditional industries” also.
A mixed picture, in other words, veering more towards mapping of potential, generally, than description of actualities.
The reason for the Report’s positive assessment of China’s AI prospects is that “its huge population can generate a tremendous volume of data, which is a prerequisite for ‘training’ AI systems” — these systems “must be fed huge quantities of data so that they can ‘train’ themselves and continuously improve and refine their output”. Also that its “wide range of industries provide a fertile market for deployment”.
In terms of the likely overall impact (of AI enabled automation), the biggest is foreseen to be on employment, alongside the potent “productivity injection” that would be imparted to the economy.
“Hundreds of millions of … jobs made up of routine work activities and predictable, programmable tasks will be particularly vulnerable. While impact on the labour market is likely to be gradual at the aggregate level, it can be sudden and dramatic at the level of specific work activities, rendering some jobs obsolete fairly quickly. Overall, AI will raise the premium placed on digital skills while reducing demand for medium and low skilled workers…”.
The Report recommends five priorities for China’s AI strategy:
“building a robust data ecosystem; spurring adoption of AI within traditional industries; strengthening the pipeline of specialised AI talent; ensuring that education and training systems are up to the challenge; and establishing an ethical and legal consensus among Chinese citizens and in the global community”.
India
All five apply with equal validity for India, muted is mutandis i.e. after making allowance for the different, lower, degree of preparedness of the economy at the ground level for making the AI transition.
So also the other offerings that are made to China in the Report in respect of “the three key building blocks of AI development: data, algorithms and computing power”.
As such, the contents of the Report provide a useful point of departure for preparing for the advent of AI. One hopes that strategic analysts and senior decision-makers in NITI Aayog (and in the line Ministries concerned such as the DOE, DIPP and MOD) would be at it already, with expert and inter-Ministerial Task Forces (and a nodal point) mandated to identify both thrust areas and weaknesses for accelerated/remedial action.
Strategic Considerations
Unlike the First and Second Industrial Revolutions (which passed India by) and also unlike the Third (digital one, which the country stumbled into somewhat fortuitously thanks to Y2K), conditions would seem to be conducive for India to attempt to be “ahead of the curve” in this Fourth one on AI, robotics, nano-technology, etc, and usher it in pro-actively. (Instead of waiting passively for it “to happen’ via a market driven trickle down.)
What is needed is a strategic approach — a “central mind” of Government, as it were, focussing on the different dimensions in all their complexity, so that timely decisions to do whatever it takes to join the front-ranking countries in utilisation of AI enabled systems all round, in manufacturing above all, at the earliest can be taken in a holistic and long term perspective.
The IT revolution in the country has already brought in some early harvest of incipient AI aspects (in the form of Apps, e-commerce, etc). It has also provided the requisite base for an AI take-off. But our sights in this regard should not be limited to the digital domain. Howsoever advanced (in “intelligence”) the virtual world might get, it is the real, brick and mortar economy that forms the mainstay of human existence and which should therefore be the focus of the national endeavour to tap AI. AI-led manufacturing in particular, because of its ubiquitous (and unimaginably huge), impacts on costs and quality of almost every kind of produce.
All forms of contests between nations, leading to the ultimate one in “comprehensive national power” (to borrow a Chinese, albeit commonsensical, concept), will boil down to a question of cost (and quality) competitiveness in manufacturing — a preview of which is already evident in more and more countries throwing in the towel before the gigantic (and seemingly invincible) sweat factories in China.
And manufacturing not being a strong point of the Indian economy, here lies the rationale for self-conscious “strategising” to foster “AIsation” (instead of leaving it to market forces to do at their own pace). A laissez-faire approach is fine in the digital domain, given the vibrant scene in the virtual world, but not in the real one. (Except perhaps in selected FMCG and other luxury goods sectors that are on “auto-pilot” course, and would be best left undisturbed.)
For the rest — for the bulk of the manufacturing/mining sectors — a strategy for “leap-frogging” on the strength of AI-led systems will have to be devised. This is the only way for late industrialisers like India to avoid getting caught in a bind (because of climate change begotten limits on per capita consumption of energy and raw materials) in their effort to catch up with the affluent economies.
The strategic approach thus entails, necessarily, going all out to think, and act, big. And afresh, out of the conventional wisdom box. And even naively, may be, but stemming from national needs and priorities.
The Vision Thing
Two possible components of a national “AIsation strategy” offered in that vein for consideration.
First, could collaboration in AI enabled manufacturing related technologies — of any kind (not excluding outright purchase), in any sector, available anywhere — (be made to) form part of the “wish list” of the nation’s diplomacy in its interaction with the industrially developed worldIJ Appropriate desiderata in respect of each such country would need to be identified. This is a painstaking and time-consuming (inter-Ministerial) task that can obviously not be done at the last minute, on the eve of a VVIP visit.
Second, how about declaring priority for AI-enabled robotisation of tasks/activities that human beings ought to be spared from doing on normative grounds — of human dignity, health and safetyIJ Segregation of plastics and glass from urban waste; handling of hazardous hospital waste; clearing of sludge in sewage pipes, dredging of filthy “storm water” drains and frothy lakes; skinning of hides from carcasses of dead animals and manual scavenging being some obvious examples.
Could that prioritisation be given the form of dedicated “National Technological Missions” mandated to work on, and gather together all necessary technologies and “know-how” for, making/assembling suitable robotsIJ With outright import of “demonstration” pieces in the interregnum until domestic manufacturing capability is developed, irrespective of costIJ The pilot “demonstration” projects would serve to spur those engaged in development of indigenous production capacity to deliver faster. At the same time, they would manifest public reaffirmation of the State’s commitment to a scientific approach to socio-economic problems in favour of the downtrodden and the ‘outcast(e)’ — antyodaya in some tangible way — and thereby go some distance in nurturing a modern sensibility in society in the process.
(The writer, a former Ambassador, is a strategic analyst with a background on China. He is currently an Adjunct Professor at the National Institute of Advanced Studies, Bengaluru where he anchors the “India in the Chinese Media” project (www.niasindiainchina.in))