Intelligent Data Centres Issue 19 | Page 41

EXPERT OPINION Roadblocks to success Frequently, companies leading significant AI research and development do so without important initial input from IT. As a result, teams unfortunately produce shadow AI – AI infrastructure created under IT’s radar, which is challenging to successfully operationalise and ultimately unproductive. Companies can avoid shadow AI by leading with an infrastructure strategy specifically optimised for AI. The survey highlighted incalculable costs as the top issue (21%). From the need for new investment in people and equipment, to unforeseen costs along the winding road between AI design and deployment, to rapid innovation and shifting of technology needs, AI investment is potentially massive and difficult to predict. Moreover, internal disconnect between IT and AI development can result in low ROI if the company fails to deploy the technology. The lack of in-house expert staff also poses a significant challenge. Companies typically need to bring on specialised developers, which can be costly and requires time for new staff to learn the business to meet AI design with organisational goals. Inadequate IT equipment also blocks companies from envisioning how AI fits into their operation. According to the survey, many worry their current infrastructure is not optimised to support AI and fear data centres have reached full capacity. Roadblocks at the strategy phase are largely similar across industries but DESIGNING AI INFRASTRUCTURE IS COMPLEX AND OVERWHELMING AND, AS A RESULT, 76% OF BUSINESSES REGARD INFRASTRUCTURE AS AN OBSTACLE TO AI SUCCESS. www.intelligentdatacentres.com Issue 19 41