I was excited to learn from a recent BCG survey how many South Africans are open to exploring the benefits of AI, with almost a third of those who are aware of the technology already using it. It’s encouraging to know that across the country, there’s significant curiosity and enthusiasm for AI. And with the government taking key steps in the right direction and inviting feedback from industry experts on its new National AI Policy Framework, we have a great opportunity to help shape the country’s AI development. The challenge, for stakeholders across industries, is to come together to make AI easier to use and quicker to adopt. This way, we can fast-track a shift in our transformation trajectory, altering our course from where we are now to where AI is moving in the future.
For AI innovation in South Africa to be meaningful, it must drive considerable economic growth. With analysts predicting that the technology could expand Africa’s economy by as much as 50% of current GDP by 2030, the potential is significant. But what’s crucial here is making sure every business can tap into the AI opportunity, beginning with ensuring they have the right digital infrastructure in place.
Key considerations to accelerate AI innovation
When it comes to getting the most out of AI, businesses in South Africa and around the world are finding that their current IT setups are insufficient. With data scattered across public clouds, private clouds, on-site and at the edge, the resulting silos make it tough to achieve the real-time data access that AI applications thrive on.
Moreover, when workloads aren’t placed on the optimal compute platform—whether that’s in the cloud or on-site—efficiencies take a hit and AI outcomes suffer. Interestingly, many enterprises around the world still stick to using only private clouds (43%) or public clouds (12%), while others solely use on-premise data centres. They miss out on the benefits of spreading workloads across different environments. These were the results from a recent survey of global IT leaders, commissioned by Hewlett Packard Enterprise and conducted by Sapio Research.
It’s perhaps not surprising that the survey revealed data management as a top priority for AI success. But here’s the challenge: most enterprises are overly confident about how ready their data is to be used for AI applications. This is especially true for those aiming to roll out AI projects at scale. While 72 percent of organisations surveyed by Sapio believe their data is AI-ready, a closer look tells a different story. For example, just 7% of the enterprises confirmed that they can run real-time data pushes or pulls and only 26% have the appropriate governance for enabling advanced analytics – both essential capabilities for AI projects.
Data maturity levels are still generally low, with data often being siloed, hidden, or mislabeled. This can significantly alter the results provided by AI applications, that depend on reliable data. No matter where a company is on its AI journey, the success of any AI initiative hinges on having solid data integrity, easy access, reliable storage, and good governance.
This brings various infrastructure enablers into play, like network-as-a-service for better connectivity, and makes it pivotal for enterprises to leverage hybrid cloud infrastructure architecture. It allows data to be stored anywhere but still treated as one unified, reliable source.
Luckily, many IT leaders are already seeing the benefits of moving to a hybrid cloud setup, as the research indicates. However, what they might not fully grasp yet is the importance of adopting a cloud model that’s ‘hybrid by design’.
How ‘hybrid by design’ can enable AI success
So, what is hybrid by design? Simply put, it’s a cloud operating model that unifies all IT environments by managing them through a single hybrid cloud platform. In the context of AI, where workloads may run in public clouds, private clouds, and on-premise data centres, infrastructure unification is critical for gaining the performance, scalability, and flexibility required for success.
What does this mean in practice? The ‘hybrid by design’ model makes it easier to manage different kinds of IT infrastructure and make them work together. It also means that IT can apply the same rules and security measures to all of these different infrastructures. This way, users, devices, and applications can access data from any environment, and get a consistent cloud experience.
In simplifying the management of multiple infrastructure environments, ‘hybrid by design’ also provides IT staff with the required visibility that enables effective control of costs and data. It supplies the capabilities required to meet business goals. For example, when scaling AI, the goals could be gaining business insights, speeding innovation, and enhancing sustainability.
From data unification to compute flexibility and optimised workload placement, the ‘hybrid by design’ strategy offers innovative enterprises numerous significant benefits for appropriately supporting AI initiatives, ultimately providing them with an IT infrastructure approach that can accelerate their AI agenda.
How can we advance and future-proof AI transformation in South Africa? We need to start by making the most of our existing IT infrastructure and finding ways of simplifying infrastructure updates where we don’t yet have what we need. South Africa’s National AI Policy Framework is a big step forward. But we can’t stop there. We need IT solutions that can help us reach our national goals and make the most of AI. A ‘hybrid by design’ model can do that. It can give us the best way to use our existing IT infrastructure and help enterprises establish a clear path to unlocking AI’s vast potential.