Reviews
5 years ago

AI strategy: Addressing country-specific issues

Published :

Updated :

As scope for application of artificially intelligent (AI) machines in most sectors is expanding, the threat of job losses in these sectors is gradually intensifying. This is why, there has been a call for formulation of national AI strategies in different parts of the world. Already countries like China and France have come up with their separate national AI strategies. According to a media report, till the end of 2018, 23 countries had unveiled their AI strategies. Although a bit late at creating an AI initiative, the USA finally launched a national strategy on AI in February 2019 with the signing of an executive order by President Donald Trump.

Despite the likelihood that developing countries are going to get marginalised more than their advanced counterparts, most of the developing countries are yet to come up with their national strategies. In formulating the strategies, should the developing countries follow advanced countries as global best practices? Is there a substantial difference between scenarios prevailing in different countries, which are likely to be affected by AI? Do the goals vary, demanding unique country-specific strategic response?

At the outset, AI opens the possibility of delegating cognitive roles from humans to machine in productive activities. Invariably, such role delegation leads to job loss. Often such cognitive role delegation leads to human-free production, as already machines are getting better in adding energy and performing manipulation task over human workers in production sector. As a result, the scope of re-skilling of workers, primarily engaged in labour-intensive manufacturing, for making them eligible for higher value addition jobs is narrowing down. Such a likely reality appears to be a serious threat for the economic growth of many developing countries, including Bangladesh. Still, the potential of AI appears to be a blessing for some countries, particularly for OECD (Organisation of Economic Cooperation and Development) members, and China. Due to aging populations and low birth rates, these countries are looking for machines as an alternative to humans in productive activities, be it for producing goods or services. Moreover, with the rich research and development (R&D) infrastructure and earlier demands for AI innovations, start-ups pursuing AI innovations are increasing the opportunity for high-income jobs in advanced countries.  

Developing countries, so far, are relying on labour for adding value to the global value chain of production. If AI innovation makes machines better as well as cheaper than labour, developing countries run the risk of suffering from the erosion of wealth-creation ability.

On the other hand, most of the developing countries, except China, have very high youth density. How can these countries create jobs for a growing number of job-seekers, in an age when machines are a better alternative? In contrary to the past, when jobs were migrating from high-wage countries to labour surplus less developed countries, AI encourages innovation to create jobs in developed countries, consequentially killing jobs in developing ones. Therefore, by copying the AI strategy of developed countries as the best practice, developing countries will be intensifying their sufferings further. 

Technology is providing a better alternative to the low-cost labour through AI innovations. As a result, developing countries are under pressure to adopt AI-intensive production machinery, often imported, for maintaining their global competitiveness, consequentlly eroding labour advantages and killing jobs. But if they do not do so to protect jobs, the economy slows down as both industrial exports and import substitutions in quality and cost suffer. On the other hand, China and OECD member countries gain from increased deployment of robotics and automation to address rising labour shortage and increasing wages. As a matter of fact, AI has a natural tendency to create few high-paying jobs in innovation, resultantly killing a larger number of production-related jobs. In developing countries, the AI strategy should reduce the threat of impending job loss while ensuring economic growth. How to have job-filled growth in the age of AI- should be the core strategic component for developing countries. 

On the contrary, software-intensive AI machines have a high scale, scope, and network externality effects. Most of the AI innovation start-ups require support, sometimes over a decade or so, to reach profitability. As a result, countries having strong R&D capacity and risk capital base can continue to acquire AI-centric competence to be better and cheaper at monopolising the global market. There has been an argument that like many other technologies in the past, AI has the potential to create new wealth. A global consultancy has predicted that new wealth created by AI can be as much as $13 trillion by 2030. Then why is there a concern? Unfortunately, the market economy will not distribute that wealth uniformly. As AI is likely to reduce the market value of labour, increase the demand of high-end innovation competence, and intensify monopolistic market power accumulation (both in local and global contexts), there is a growing risk of AI fuelled increasing inequality between individuals, firms, and countries. Therefore, the formulation of strategy in coping with as well as leveraging AI should be taken under consideration.

As component technologies surrounding AI are easily accessible, there could be an argument for scaling up investment in education, research, publications, patents and risk capital finance to support innovation and start-ups, in order to leverage the possibilities of AI. Despite the potential, most AI innovations suffer losses for a prolonged period of time, often requiring R&D finance and operational subsidy for decades.

Additionally, in any AI-innovation space, the strategy of monopolising the market often in the global context appears to be the winning formula. Despite the strong relevance of intellectual assets, acquired through education and R&D investment to pursue AI innovations, there has been no natural correlation between conventional indicators and success in AI innovations. As such, returns on investment in case of most AI innovations are risky. For example, China's staggering investment in R&D and record numbers in publications and patents are failing to produce a proportionate economic return. Without taking into consideration the challenges to succeed in generating profitable revenue out of AI innovations, often developing countries run the risk of slowing down the economy by increasing investment to pursue AI innovations.

Citing the phrase "do not reinvent the wheel", often it is preached that developing countries should adopt best practices pursued by advanced countries. But this does not apply to AI and varying implications on jobs and growth in different countries. So, developing countries should carefully analyse their situations, draw a lesson from others, and formulate AI strategy within sound theories of technology, innovation, and functioning of the market. Otherwise, by copying the strategy of advanced countries, along with a weak focus on the functioning of the market being affected by increasing monopolistic market power accumulation and nature of AI innovations and start-ups in developed countries, developing countries like Bangladesh run the risk of losing both jobs and face economic slow down and increasing inequality. 

M Rokonuzzaman Ph.D is academic and researcher on technology, innovation and policy.

[email protected]

 

 

 

Share this news