Case studies: AI in developing countries
In our first article exploring this topic, we looked at the rationale behind why governments should care about artificial intelligence along with the particularly profound effect that the technology will have on poorer nations. That being said, in this piece, I aim to explore some examples of AI policy that have been formulated by governments in developing countries worldwide. For transparency purposes, we’ll be working with the classifications that have been developed by the UN’s World Economic Situation and Prospects 2019 report.
South America: Peru
Figure 1: The fall in global oil, gas and mineral prices over the past decade have exposed South America’s overdependence on commodities, causing sluggish economic progress across the board.
Against the backdrop of falling productivity and economic growth, adopting artificial intelligence in the right manner may be a step in the right direction for South America. In an analysis of five economies in the region, Accenture estimates that AI can add up to an entire percentage point to the continent’s annual economic growth by 2035.
For Peru, one of the fastest growing economies in South America, the report also proposes that the technology can add US$43 billion to the nation’s GVA by this time. The added value within the country’s wholesale and retail sectors alone from implementing the technology is projected to be a whopping 17% of the above figure. Like many of the nations we’ll consider, challenges such as talent scarcity, inadequate educational infrastructure and a weak culture of research will make it difficult for Peru to even begin to realise these gains.
Despite this, the discussion surrounding AI amongst South American policymakers seems to be lacking, and this appears to be a symptom of a larger problem. As a country economically driven by mining, policymakers are yet to realise the value of fostering a STEM-driven economy, let alone artificial intelligence. As of 2018, the government invested only 0.08% in innovation-based initiatives, which is far below that of its neighbours.
Sub-Saharan Africa: Kenya
Over the past several years, AI development in Africa has blossomed. Several groups and events focused on practicing and dispersing information about the technology have been established, such as the annual Data Science Africa workshop and the organisation Deep Learning Indaba. The continent has also been the subject of major firms’ AI expansion efforts. IBM Research has opened an office in Nairobi as well as Johannesburg. Similarly, earlier this year, Google established a new AI lab in Accra, Ghana.
More specifically, within Kenya, groups such as AI Kenya have formed with the intention of “bringing together like-minded people to discuss, build and grow the Artificial Intelligence and Data Science ecosystem in Kenya and East Africa”.
In February 2018, the Kenyan government announced the Blockchain & Artificial Intelligence task force, consisting of 11 members from academia and industry. The group’s first goal was to provide recommendations to the government regarding the harnessing of emerging technologies over the next half-decade. In July this year, the group released their long-awaited first recommendations in the “Blockchain Taskforce Report”, which was submitted to the Ministry of Information and Communications.
Kenya is the only sub-Saharan African state (out of the 46) with an AI task force that is working towards a national strategy. As a leader in this area, they carry the weight of designing a framework that can succeed in the face of challenges unique to most of Africa, including but not limited to: major infrastructural challenges, poor regulatory conditions, gaps in the skilled labour market.
It seems that whilst conversation surrounding AI has been slow in Sub-Saharan African governments, the private sector and the general public have taken notice of the emerging global trend. The development of the technology in Africa as a whole, it seems, may revive the original promise of AI: to create technology that can help to solve the world’s most pressing social issues. One such example is Google’s work with farmers in rural Tanzania; a team of researchers created a machine learning model that could diagnose early stages of disease in the cassava plant, a staple crop in the region. What’s most exciting is that the unique challenges that the African environment poses may in fact reveal new avenues for AI to explore.
Southeast Asia: Malaysia
Earlier this year, in partnership with the International Data Corporation, Microsoft conducted a study on artificial intelligence in the Malaysian economy. In a survey of 100 business leaders and 100 workers, it is believed that by 2021, AI will allow the rate of innovation to almost double and increase employee productivity by 60%. Yet, only 26% of the organisations in Malaysia have adopted some form of AI.
In October 2017, Malaysia’s former Prime Minister, Najib Razak, announced plans to develop a National Artificial Intelligence Framework, which would serve as an expansion of the National Big Data Analytics Framework. Since the election of the current Prime Minister, Mahathir Bin Mohamad, in the May of 2018, little information has been released, but it has recently been revealed that the Malaysia Digital Economy Corporation (MDEC) are expecting to complete the framework by the end of 2019.
The much broader-viewed Digital Transformation Acceleration Programme for medium and large companies was established in April 2018, which provides structured lab sessions for firms to: “identify pain points and opportunities in the digital space, uncover potential solutions and implement proof of concept or minimum viable product with measurable outcomes.” This is also run by the MDEC.
Middle East: Saudi Arabia
As a nation that consistently reserves a spot in the world’s top 20 economies by GDP, perhaps it’s to no surprise that Saudi Arabia’s measures to adopt AI have been far more comprehensive.
Dr. Ahmed Al Theneyan (Saudi Arabia’s Deputy Minister of Technology, Industry and Digital Capabilities) had the following to say on the topic of Saudi Arabia’s Vision 2030 development program:
“Saudi Arabia is undertaking the largest and most ambitious economic reform and transformation program in its history...Digitization and artificial intelligence (AI) are key enablers of these wide-ranging reforms.”
Financially, the government has pledged investments in AI through the Public Investment Fund, a sovereign wealth fund. Commitments to domestic initiatives by the government and private sector have also been made, spanning from the construction of the “smart city” Neom, to becoming the first country to grant citizenship to a robot. The government has even deployed Sophia as a representative for the kingdom at events, most notably at the 2018 UN Summit on “AI for Good”.
It is critical to note however, that Saudi Arabia currently does not have an official national AI strategy. Instead, the government has rather indirectly demonstrated its commitment to the technology by announcing a Vision 2030 plan that requires AI to achieve. In a 2017 study, PwC’s consulting arm estimated that AI could contribute a whopping $135 billion or 12.4% to the nation’s GDP by 2030. It’s no wonder that Saudi Arabia’s ‘full-steam ahead’ attitude to adopt AI is filled with optimism.
Figure 2: Sophia is a social humanoid robot developed by Hong Kong-based company Hanson Robotics. She was granted Saudi Arabian citizenship on October 25th, 2017.
So, what now?
Undoubtedly, there are trade-offs that exist in just how far developing countries should go in terms of developing their national AI capabilities. Not only this, the pace at which this occurs should also be considered. From the above analysis, there are a few common trends that can be discerned. Primarily, most of these countries have only recently developed national strategies to address this emerging technology. It seems that while policymakers have shown varying degrees of interest, for the most part, no major financial commitment has been made.
Whether or not these nations can strike the right balance between development and regulation is anyone’s guess. What we can say however, is that any initial start is a promising sign; critically, every government must understand and acknowledge the potential for AI to transform their economies, for better or for worse.
Author: Alden Vong | Communications Officer