The Challenge of Data Management in the AI Revolution

2023-08-18

According to a new global study commissioned by S&P Global Market Intelligence and WEKA, companies that invest in data management now will become leaders in AI in the long run. Among the respondents surveyed, 69% have at least one ongoing AI project, with 28% already reaching an enterprise scale. While companies and research institutions seeking to create new value propositions are accelerating the adoption of AI, the study points out that challenges in data infrastructure and AI sustainability still hinder the successful implementation of large-scale AI. This is mainly due to the rapid development of generative AI in enterprises by 2023. AI adoption continues to increase, but enterprise scale remains a challenge. WEKA and S&P's joint research surveyed 1,500 global AI decision-makers and revealed these findings. It identified the opportunities and obstacles encountered by companies on their AI journey, as well as the unique drivers of AI adoption across various industries globally. The research also provides insights into the steps companies need to take in the future to achieve AI success. 32% of respondents identified data management as a technical barrier to AI/ML deployment. Furthermore, evidence suggests that many companies' current data architectures are unable to support the AI revolution, compared to challenges in security (26%) and computational performance (20%). 77% of respondents believe that traditional architecture and data infrastructure impact their sustainability, and a total of 74% of respondents stated that sustainability is an important or critical motivation for shifting more workloads to the public cloud. 68% of respondents also expressed concerns about the impact of AI/ML on their company's energy usage and carbon footprint. As AI projects further develop, adopting hybrid approaches and multiple deployment locations are necessary to support workload demands. According to the research by S&P and WEKA, those who run AI/ML on the public cloud are most likely to adopt hybrid approaches, while those who do not use the public cloud are less likely to do so. Liran Zvibel, Co-founder and CEO of WEKA, said, "Traditional data infrastructure has a direct and negative impact on their ability to use AI at scale efficiently and sustainably because they were not designed for modern performance-intensive workloads or hybrid cloud and edge patterns." "Just as you wouldn't expect to power modern electric vehicles like Tesla with battery technology developed in the 1990s, you can't expect data management methods designed for last century's data challenges to support next-generation applications like generative AI." "Building a modern data stack designed to support AI workloads from edge to core to cloud will make companies the leaders and disruptors of the future."

RECENT AI NEWS

RECENT AI TOOLS