Artificial Intelligence is tough to define and 'terms and conditions' do apply, but companies in Africa that want to emerge as a data-driven AI-first businesses in the smart technology era need to act fast and deep dive into their data structures, processes and procedures.

This is according to ICT professionals and business owners who attended the ITWeb AI Summit hosted in Johannesburg yesterday.

Jacques Ludik, president and founder of the Machine Intelligence Institute of Africa; as well as CEO and founder of Cortex Logic, said while he senses a definite desire in Africa to advance technically and leverage the benefits for business, there is concern that the continent could still miss the opportunity.

Ludik said he was inspired when he attended the World Economic Forum and witnessed the vision, need and determination of regions like Rwanda which stated their determination to "take on the challenge and not be left behind".

According to Ludik countries like China, Canada, France and the US have national tech-driven economic development strategies in place and continue to plough millions into these strategies. He added that the UN and organisations like the World Bank remain focused on ways to leverage exponential technology like AI and machine learning to transform economies.

"But do we have that vision in Africa? We can't be effective if we don't collaborate. We need collective wisdom, we need multiple stakeholder input ... this is one of our biggest problems. We are too isolated around this, we need to try and leapfrog.... Just look at M-Pesa for example, and what it has done for mobile money," he said.

Speakers used the analogy of getting into a car and embarking on a road trip to describe the approach a business could take in their quest for AI-based digital transformation.

'Waypoints' on this journey include understanding the problem that needs to be sorted out, the AI 'readiness' of a business, costs involved (the amount of budget that should be made available), leveraging existing infrastructure as much as possible, partner to 'piggy-back' and progress on the journey quicker, and available skills base.

However, the first point of departure is data and data management.

The first step is to have clean, reliable data in place. "Junk in, junk out, right? If you want to start addressing business problems, remember to start with generic value drivers... a lot of the end result is prescriptive actions or helping decision making. Intelligence is difficult to define. It starts with business KPIs and then you can work it back. You got to make sure data driven solutions are organised to give rapid access to all types of data. You've also got to make sure data pipelines can be reused. You need data engineers, data architects etc. to deliver end-to-end AI solutions," Ludik added.

AI is described as the new electricity and while most agree it can never replace the human factor in business, there are distinct advantages for business that can exploit this exponential technology, tape into the new economy and leverage 'plug and play'.

The reality is that those businesses that move quickly, adapt and leverage this technology that underpins the 4th industrial revolution will advance – and end up out of reach of their competitors.