Kathy Gibson reports from Gitex Africa Morocco – As generative artificial intelligence (GenAI) permeates every organisation, the way we do business and operate our companies has got to change.

African business leaders believe that GenAI will help their companies to achieve competitive advantage, with the value coming from creators using the tools to increase productivity and savings, says Saad Toma, GM of IBM Middle East and Africa.

Indeed, in a recent IBM survey, 43% of CEOs said they will use gen AI to inform strategic decisions. Meanwhile, AI is projected to enhance human productivity and unlock $16-trillion in value by 2030.

However, adopting AI is not without its obstacles, Toma points out. Globally, 82% of IT professionals say IT complexity is impeding success in deploying AI, while 55% of business leaders lack key information regarding their technology spending decisions.

In Africa, many organisations face barriers such as costs, market and regulatory factors, workforce readiness, infrastructure, skills gap, ethics and governance.

Toma believes that, to overcome these challenges, organisations must move to an AI-first approach, where AI is integrated into their business strategy across the lifecycle.

“Being AI-first enables businesses to be value creators rather than solely value users. Companies that will lead their respective industries for the next decade or two will be the ones that decide to be AI-first.”

The biggest value will be achieved in increased employee productivity, with more efficient workers able to build value quicker and better.

Toma cautions organisations to bear five fundamental truths about AI in mind when they embark on their transformation journeys.

The first is that AI must be multi-modal, employing various commercial and open source models. There are around 200 large language models (LLMs) available so organisations need to be inclusive.

A multi- hybrid-cloud environment will be key to success, so that AI models can run where the data, workflows and apps live.

Governance is crucial to ensure the models and data have provenance, and monitor for hallucination, bias and explainability.

The environment must scale in order to add value, so organisations can pick the right use cases and deployment for the best possible return on investment (ROI).

And data matters, Toma stresses. Too often pilots miss in production due to data quality, access and security issues.

IBM’s has recently added new upgrades to its GenAI platform Watsonx, including an Arabic LLM, the IBM Concert nerve centre that applies AI to the way we manage IT, the InstructLab community-driven language model development, and Open Source Granite models that deliver accurate results with fewer parameters.

This last is important for Africa, Toma points out, since access to GPUs can be complex and expensive. “If you can do inferencing at a lower cost – and fewer parameters means lower costs – then you can use lower-GPU technology to provide the same outcomes.”

Toma points to some successful case studies that IBM has delivered using GenAI on Watsonx.

The company is working with NASA on geospatial modelling of the earth, including heat modelling. This has implications for Africa in terms of finding the right land and water resources for expanded agriculture.

Natwest uses a virtual assistant Cora to offer conversational responses to complex data queries.

IBM consulting is collaborating with Riyadh Air to build a hyper-personalised digital-first guest experience across mobile, web and in-flight entertainment as the airline readies for its first flight in 2025.

Transport for London is using GenAI to keep the public moving safely, reliably and sustainability with centralised management. So far, the system has had no downtime, and is saving the organisation money.

Sund Baelt is using an AI-powered IoT (Internet of Things) solution that inspects the longest bridge in the world to monitor for faults or maintenance issues. The system is expected to extend the lifespan of the bridge by 100 years, while saving 760 000 tons of CO2. Where mountaineers were previously employed to do so, drones now examine 300 00 square meters of concrete every day. The AI system inspects the images to keep an up-to-the-minute record of the structure’s health.