The road to AI: Navigating the detours, critical gaps, and blind spots

[Sylvia_Hooks]

Ready to move full speed ahead on your AI journey? As you formulate and advance your AI strategy, get insights from your peers to drive your success. New research sheds light on how your peers are navigating the new terrain of AI innovation, and their challenges with strategy, leadership, and data management.   

I can barely go a day at the office without hearing some mention of the massively transformative power of artificial intelligence. AI has emerged in record time as an essential cornerstone of innovation and growth not just for tech companies, but in virtually every industry and for every business process.  

With the AI market poised to exceed $738 billion by 2030, according to Statistica, organizations are keen to harness its untapped potential. Yet, while enthusiasm for AI abounds and self-confidence is high, new research commissioned by HPE reveals that the path to AI success isn’t necessarily a straight line.  

The reality is that we are all grappling with the sheer newness of our different AI implementation journeys, and the struggle is real. The study, conducted by Sapio Research, surveyed 2,453 IT decision makers across 14 markets, reveals the bumps on the road to AI adoption.

Although we are in the dawn of AI’s practical application, the research showed that 44% of IT leaders believe their organizations are already fully set up to realize the benefits of AI. It’s a high figure—indicating that overconfidence can easily lead to missteps and disillusionment. So where are we most exposed along the road to AI? Let’s go back to the basics and look at the “why, who, and how” of successful AI applications. 

Why use AI: Develop a comprehensive strategy 

Let’s dive into the strategy zone first. Gaps in the “why” seem to center around an organization’s ability to build an effective AI strategy, including establishing measurable goals and outcomes. The research shows that despite the widespread excitement around AI's potential, many organizations struggle to articulate a clear and comprehensive AI strategy that will deliver measurable business value.  

While many organizations have established official AI strategies and set ambitious goals, there's a noteworthy lack of cohesion and integration, which will slow progress. Only a fraction of organizations has adopted a best practice, end-to-end approach to AI. Disparate strategies and goals for individual business functions only lead to fragmentation and dilution of AI initiatives.  

Another strategic area that has room for improvement: sustainability. We know that AI workloads are massively resource intensive. In the survey, 92% of IT leaders indicated that their organization has actively taken steps to reduce the energy consumption of their AI applications. But upon closer examination, only 44% say they are actually monitoring their AI-related energy consumption. Clearly, we have more work to do to meet sustainability goals in tandem with disruptive innovation. 

Who’s driving? Get all your AI champions in a room 

Another gap we see emerging is in the “who” category, as businesses try to enlist the right decision-makers to champion their AI initiatives. While 97% of IT leaders believe the right people are currently involved in their AI strategy conversations, there's a disconnect between decision-makers and influencers. Decisions around AI strategy warrant a broad spectrum of input, including from C-suite execs, IT leaders, AI/ML engineers, data scientists, and data center architects, to name just a few teammates on the roster.  

While IT leaders express confidence in the involvement of other stakeholders, today’s decision-making around AI strategy primarily falls within IT departments. This imbalance underscores the need for closer collaboration between the C-suite and IT leaders to ensure strategic alignment and effective decision-making around a company’s AI ambitions. 

How to succeed: Build and scale the AI infrastructure  

The third gap surfaces in the “how”—the prepping, managing, and secure handling of the massive volumes of data required for AI models. We all now agree that data is the priceless asset of our time that drives innovation, but the AI output is only as good as the data input. While investments in AI infrastructure and software are on the rise, organizations will need to integrate AI into business processes across the organization, whether to support customer experience, IT systems, manufacturing process automation, or other business functions. Data management has the potential to become a critical bottleneck, with low data maturity levels hindering AI model development.   

Where the data infrastructure required to build and scale AI-driven apps lives is critical. Gen AI apps are creating tremendous excitement among consumers and the media, but with AI offering such a strategic advantage, many organizations will develop and run AI apps on private infrastructure. In fact, the research shows that 80% of organizations are currently running AI models on infrastructure they manage. Public clouds have offered a great space for experimenting with small-scale, low-risk projects—but they don’t have a place in larger deployments.  

The research also sheds light on a disconnect between perceived readiness and actual understanding of networking, storage, and compute requirements of AI In the survey. Fewer than 6 in 10 respondents said their organization is completely capable of handling any of the key stages of data preparation—accessing, storing/protecting, processing, and recovering—used in AI models. This gap highlights the need for a more nuanced approach to AI enablement, as businesses should take the time to better understand and address AI’s resource-intensive requirements. 

And piling onto the data challenge are the knotty issues of ethics and compliance. The survey shows the generally held views that legal, or compliance and ethics, are the least critical elements for AI success—selected by only 13% and 11% of IT leaders respectively. This perception is a serious weakness, as these elements are already becoming increasingly important in both consumers’ eyes and for regulatory compliance.  

Slowing down to mind the gap 

No doubt about it, AI is most certainly here to stay, despite the gaps we're seeing. So, what's our next move? If we want to really tap into AI's full power, we've got to aim for a smoother approach, from start to finish. And HPE has some suggestions for this: 

• Don’t rush. Begin with a clear purpose, defining business outcomes and involving leadership and knowledge holders from across your organization. Craft an AI strategy that aligns with your overarching goals. 
• Embrace a holistic strategy. Develop an overarching AI strategy that brings together the big three: ethical considerations, business goals, and sustainability targets—and remember that you need to monitor AI-related energy consumption. 
• Foster collaborative leadership. Assemble your dream team of AI players. Encourage strong teamwork among top-level executives and IT leaders, so they can blend business smarts with technical know-how to push forward your AI innovation.  
• Invest in enablement. Focus on investing in data management as well as the network, compute, and storage infrastructure either in your private data centers or with your cloud partners to support AI innovation at every stage across the entire lifecycle. 

 By tackling these important areas and avoiding the detours and potholes, you can set up your organization for long-term success in a world that (before we blink) will be driven by AI.