TECHNOLOGY STRATEGY
moved workloads back from cloud to on-premises infrastructure.“ I have seen transformation fail. I have seen organisations backing out as well,” Syed admits.“ But again, they have to stand up with a new and different strategy because they cannot always be sitting on that old legacy world.”
Innovation at the edge One of the most significant trends emerging in 2025 is the shift towards edge computing and distributed architectures. AI-driven migration tools are growing faster than anything else, with 28 % annual growth in the last couple of years, while edge-to-cloud architectures are rising fast: 1 % in 2023, accelerating to 25 % in 2024.
John sees this as essential for the next phase of cloud evolution, particularly as AI workloads become more demanding.“ If you’ re trying to calculate the curvature of space-time of an object approaching the event horizon of a black hole, or you want to know what trainers look cool, you probably don’ t want the same amount of information in there,” he quips, referencing the movement from large language models( LLMs) to smaller, more focused models( SLMs).
This shift towards distributed computing is being driven by both cost and performance considerations.“ Narrowing it down means you can move it much closer to your customers, your end users of those services. They get a better experience, you get a better experience as the service provider because your customers are more engaged,” John explains.
The bottom line Looking ahead to the next five years, the question remains whether organisations will master the economics of cloud and AI, or continue struggling with cost management and value extraction. John remains cautiously optimistic but emphasises long-term thinking.
“ You must keep a long-term eye on things, and none of this is easy. It wouldn’ t be fun if it was easy, to be honest,” he notes.“ But if you can front-load that with today’ s wages and developer time and costs, you will be so much better off in the next five years.”
94 September 2025