Why Nigeria should not be thinking of an AI Agency…yet

Earlier last week, the news broke out that the FG intends to set up an #AI agency that will be in charge of coordinating R & D of Artificial Intelligence in Nigeria. For more on this: https://t.co/XfT4tm3oN6

Although this move is in line of global trends, especially outside Africa, I believe it is too little, too soon. If the FG has decided to start considering AI, a government agency is not the first the beginning point. Other items should be in place. In this article, I'll be discussing what else Nigeria should be talking about before setting up an #AI agency, why it isn't necessary to set up an agency, and why it might fail in Nigeria if set up. Let's start.

In the global list of the top 100 institutions involved in Artificial Intelligence, only 3 of them are government agencies. Eighty-seven of them are universities, 8 are research institutions, and 2 are private enterprises. So even Google and co. are "minorities" in the global AI race. What does this tell us? The drive to build #AI should not start from the govt., it should come from the universities. In China, academics hold 52% of AI patents.

Nigeria can't be talking about AI when most of our universities are still deficient in essential tech R & D, not to speak of advanced technologies like AI.

Even amongst the top 10 AI government agencies in the world, none of them is an AI agency. They are:

- US Department of Defence

- US Department of Energy

- NASA (National Space Agency)

- National Institute of Health

- US Navy

- US Army

- US Airforce

- Jet Propulsion Laboratory

- US Army Corps of Engineers

- Los Alamos National Laboratory

The above list shows us you don't need to set up a dedicated agency before engaging in AI. Existing MDAs can participate in it, based on their needs.

You may have noticed that most of those agencies are US-based and military-inclined. What does that tell us? Defense, space exploration, and healthcare are the top priorities of the USA in applying AI. Why does Nigeria want to use AI? What do we want to use it to do?

Out of the numerous applications for AI, Nigeria needs to decide what it will focus on, based on our socio-economic realities. India focuses on using AI to solve challenges in the energy, telecoms, and other core computer science fields. China applies AI in energy management, robotics, amongst others.

Every country engaged in AI actively focuses on a set of things. Nigeria should first decide which of the numerous challenges it wants to apply AI. Why can't an AI agency do this for us, you might wonder? Because like i said earlier, there's a whole lot more doing AI than setting up another bureaucracy around it.

We need to decide what we want first from AI; what does it mean to us - the conclusions we reach will guide us on how to chart a course in AI. They are several examples of this approach.

Several countries have discussed their view on AI, its usefulness to them, and how they intend to get involved in it. Their discussions are in public documents that I'll mention

USA

- National AI R&D Strategic Plan 2016 (Updated 2019) https://t.co/A1gNdnf0Ez

- AI, Automation, and the Economy (2016) https://t.co/ahkacbQYJZ

- Preparing for the Future of Artificial Intelligence (2016) https://t.co/bsXg8KlMrX

- AI White Paper (2017)

EU

- Strategic Research Agenda for Robotics in Europe 2014 - 2020 https://t.co/l1Vt6YhnNt

- Robotics 2020 Multi-Annual Roadmap (2016)

- Gauging the Future of EU Research & Innovation (2017)

- Civil Law Rules on Robotics 2017

Japan

- Japan Revitalisation Strategy 2016

- AI Technology Strategy: Report of Strategic Council of AI Technology 2017

There are several other countries like the UK, China, France, and Germany, with separate plans and policies for AI. We can't jump the gun from planning to implementation by setting up AI. In my recently published book, "AI & Law: A Nigerian Perspective," I recommended that Nigeria set up an AI agency. But before that, I made it clear that we should first set up an ethical and legal framework, deciding how and why would be using AI and for what purposes. You can get the book for free here https://t.co/TRGbk59IS5

In summary, I've tried to prove that Nigeria isn't ready for an agency.

We need to:

a. fix our universities

b. Decide what we want to use #AI for

c. develop clear plans and policies towards achieving those ends

e. decide whether setting up a central agency is the way to go