Two interlocking technologies formed the basis of the latest Tech Monitor executive roundtable event, put together in partnership with Lenovo and Intel. In late September, senior technologists convened in central London to discuss the merits of artificial intelligence (AI) and hybrid cloud, both independently and working together.
AI – particularly generative AI (GenAI) – has the potential to transform how businesses operate, unlocking new levels of productivity, automation, and insight. At the same time, hybrid cloud is emerging as the foundational infrastructure needed to support these compute-intensive workloads. It promises the scalability of public cloud alongside the control and performance of on-premise environments. But while the potential is clear, the path forward isn’t always straightforward. From cultural resistance and skills gaps to cost pressures, regulatory demands, and legacy constraints — the challenges are real.
Cloud adoption? It’s a long story
To understand the emergence of hybrid cloud, first you must recall the two-decade tale of cloud adoption. Go back twenty years or so, and the promise of a scalable, pay-as-you-go infrastructure solution that was quick to set up and avoided expensive capital costs proved so attractive that many organisations committed to an “everything to the cloud” approach.
Only later did businesses appreciate some of the challenges such an approach presented, including excessive latency, escalating costs, and concerns about data residency. On the last of these challenges, recent regulatory changes in the United States – notably the US Cloud Act, which allows for the possibility of US authorities accessing personal data stored in overseas data centres without prior approval – have focused minds. UK-based firms are concluding that sovereignty means not just assuring that cloud data centres are regionally located but also that those who process and manage the data work within the laws of the jurisdiction, too.
The merits of hybrid cloud
The choice between on-premises and cloud, said one voice around the table, comes down to three factors: “cost, risk, and time”. Weighing up all three, there is an identifiable trend of repatriation where workloads initially spun into the cloud are now returning to on-premise. But, as one voice cautioned, repatriation shouldn’t be confused with the demise of public cloud. One only needs to look at the ever-growing profits of the hyperscalers – AWS, Microsoft Azure, and Google Cloud, among them – to note the ongoing health of public cloud. Hybrid cloud – mixing private provision with public cloud across a single interface – has emerged as the pragmatic solution.
On cloud costs, one speaker around the table shared a cautionary tale. He recalled the story of a senior technologist who had fully committed his firm to the cloud, overseeing 100% of workloads switching to virtual infrastructure. Rather than expressing pride in his achievement, the IT leader was instead exploring ways to reverse his decision. At least partially. Why? Because once made aware of a “cloud-only” approach, which would see IT costs rise every year, his CFO pointed out that this would necessitate that the business grow every year, too, without fail. He could give no guarantee and had no desire to commit to these spending increases. So while the cloud-only move made perfect sense from a technical point of view, from a financial point of view, it looked a little different.
Balancing AI workloads
AI provides the perfect prism to look at the merits of hybrid cloud. For the most part, the public cloud provides an excellent environment for experimentation. Working with relatively small data sets, businesses can spin up proofs of concept at speed before deciding whether to pursue an AI initiative. The move from experimentation to deployment often sees organisations switching from cloud to on-premise for greater control of costs and of the data under their supervision.
Those who take this path reflect the thinking behind “The Trillion Dollar Paradox”, a term coined by the venture capital group Andreesen Horowitz in 2021 to describe the contraction in cloud computing. In words that capture the essence of the paradox, authors Sarah Wang and Martin Casado wrote: “You’re crazy if you don’t start in the cloud; you’re crazy if you stay on it.”
Here come hybrid AI
During his opening remarks, Lenovo’s Tikiri Wanduragala drew a direct parallel between hybrid cloud and AI, coining the notion of hybrid AI. According to Wanduragala – Technology Leader UKI for the Infrastructure Solutions Group (ISG) at Lenovo – three flavours of artificial intelligence will emerge. First, he said, there are going to be some elements of AI that are automatically available, that are embedded in your device. Think of a translation app, for example. Second, there will be the applications provided by third-party providers that will include AI-driven functionality. Some of these apps will be new, many will be based on pre-existing solutions, and others will be developed in partnership with the customer.
Finally, there is the “do it yourself” model, where organisations will build their own, bespoke AI applications that fill an unmet need and/or provide a competitive advantage. Organisations will choose the AI models that best suit them, most likely in combination. “If you want to know what’s going to happen in AI,” said Wanduragala, “think of the cloud story. It’s going to feel very, very similar.”
Watch the video above to learn more from Lenovo’s Tikiri Wanduragala, and to hear how AI adoption is changing enterprise IT.
‘How AI and Hybrid Cloud are Reshaping Enterprise IT’ – a Tech Monitor executive roundtable in association with Lenovo – took place on Tuesday, 23 September at the South Place Hotel, London
