Artificial Intelligence (AI) offers opportunities for a developing country like Ghana to transform its economy. AI can provide solutions in all sectors of the economy. In agriculture, AI offers solutions for the detection of pests and diseases. In health, AI can help in the diagnosis of diseases using fewer doctors. AI offers enormous potential in education as the developing world grapples with issues around online education. Furthermore, AI can help detect faults in our utility infrastructure and the detection of fraud. It is, therefore, not surprising that earlier in 2020, the Vice President of Ghana announced that the Electricity Company of Ghana is piloting an AI system to detect the theft of power.
However, there are several challenges that stand in the way of a country like Ghana in our quest to adopt AI systems. One of these challenges is the lack of technical know-how. When you look at the range of digital skills required to thrive in the fourth industrial revolution, AI is located at the high-level skills end. Currently, there are no AI courses in our colleges and universities. There is also a low number of experts in areas like Computer Science and Engineering with high-level AI skills. There must be a policy that will support digital learning and training for the development and use of AI. This must include support for private training organizations, professional associations, innovation hubs, incubators, accelerators as this is where our youth interested in AI are acquiring skills.
In addition, we must pay attention to broadband connectivity if we want to benefit from adopting AI systems. Lack of broadband infrastructure, including ancillary infrastructure such as power especially in underserved and unserved areas, remains a significant bottleneck. Also, the majority of Ghanaians cannot afford broadband services due to the cost of the service and end-user devices such as laptops, tablets, and smartphones. For those who can afford, there are always complaints of poor quality of service which service providers and the regulator must address urgently. For rural areas, it is important to explore the low cost and scalable solutions that will be affordable. We need to review our broadband policy to emphasize Universal Service Obligations (USO) and specific connectivity targets. GIFEC (Ghana Investment Fund for Electronic Communications) as a Universal Service Fund (USF) must be positioned strategically to provide the infrastructure that will facilitate the introduction of broadband by the service providers in rural areas.
Data is the lifeblood of AI. In 2017, IBM suggested that 90% of the world’s data had been created in the last two years. This means that there is an enormous amount of data that is being created every day. This has several implications for AI with regards to data management, data privacy, and data governance. It is, therefore, important that our public and private organizations build the capacity to manage data assets. It is also essential that our organizations understand data privacy and ensure that the rights of individuals are safeguarded. Despite all the data that is being produced, most of the data are controlled by big corporations. Startups struggle to get the needed data, and this stifles innovation. It is important to support projects like FAIR Forward. “FAIR Forward – Artificial Intelligence for All” is a GIZ initiative that facilitates the provision of open, non-discriminatory and inclusive training data and open-source AI applications. Innovative data governance models must also be encouraged as they will make data available for various organizations under pre-agreed conditions.
Ghana must aspire to build AI systems that will not harm people but rather do good to mankind. These include AI systems that respect people’s goals, wishes, and privacy while guarding against any form of bias. It is crucial that AI engineers strive to avoid data bias not just because accuracy levels will be affected. They must also reflect on issues of gender, stereotypes, historical unfairness, prejudice, etc. All these will not happen automatically. We need to build on our Data Protection Law by formulating an AI policy and regulatory framework that will not stifle innovation but at the same time promote accountability, transparency, and privacy in the building of AI systems.
It is widely known that startups are a source of innovation. However, the business ecosystem in Ghana is not favourable to startups regardless of industry. The cost of capital is very high. Technology startups in developing countries have resorted to digital platforms to present their services to potential users. However, they are usually not successful in getting these users to pay for the services. Business models are required that will enable AI startups with innovative products and services to thrive.
Finally, I would like to add that if we resolve all the barriers above but do not encourage AI solutions with locally relevant content, we cannot create a demand for AI services. We need AI applications in local languages and AI solutions that consider our rural-urban divide, pay attention to gender, and the challenge of poor internet connectivity.
The author is Advisor, FAIR Forward – Artificial Intelligence for All”, GIZ and executive member of the Institute of ICT Professionals, Ghana.
For comments, contact +233274333510