The majority of business leaders and technical analysts believe that AI is an essential tool for business sustainability. Nearly 3 in 4 decision-makers believe that not investing in AI puts their business at risk of failure, according to a report by Exasol, a high-performance analytics database provider.
The Exasol report is based on research into data, analytics, and AI commissioned by Exasol from Vanson Bourne, an independent market research firm for the technology sector. The respondents for the study included over 800 senior decision-makers in IT and non-IT roles.
While business leaders are aware of the importance of AI, the technological challenges and regulatory requirements are slowing the process. Even the growing stakeholder pressure to implement AI hasn’t been enough to significantly accelerate AI adoption.
The report by Exasol investigates the current state of AI implementation and highlights some of the key challenges, opportunities, and future trends for businesses in the context of emerging technologies. 
An overwhelming majority (91%) of respondents agree that AI will be on top of the agenda for organizations in the next two years. The top reasons for this belief are the ability of AI to create new sources of revenue (50%), the evolving nature of roles and responsibilities (47%), and the rapidly growing competitiveness in the market (46%).
Despite the widespread excitement and understanding of the transformative potential of AI, the adoption rates don’t match the enthusiasm. A key obstacle is latency challenges in terms of speed of implementation for new data requirements.
Nearly half (47%) of respondents in the Exasol study shared that the time needed to adjust to changing data landscapes and integrating new data sources is a major obstacle to AI implementation. Other key challenges include slow reporting performance and increased data volumes. While almost all respondents (96%) shared that their organization used BI acceleration engines, yet a high percentage (69%) reported sluggish reporting performance.
“Our study further proves there is a significant gap between current BI tools and their output – more tools does not necessarily mean faster performance or better insights,” said Joerg Tewes, CEO of Exasol. Tewes recommends careful evaluation of the data analytics stack to ensure optimum productivity, speed, flexibility, and cost-effectiveness.
Another study by Vanson Bourne for Fivetran, a global leader in data movement, shows that while companies push to adopt AI, they are losing hundreds of millions every year due to underperforming AI models. There is more time spent on preparing data, and then actually building models with it.
The report, which was conducted by surveying 550 respondents, reveals that companies lose on average 6% of their global annual revenues, or $406 million, based on data from organizations with an average global annual revenue of $5.6 billion.
The underperforming AI models are built using inaccurate or low-quality data, leading to misinformed decision-making. According to the report, organizations in the U.S. suffer inaccuracies and hallucinations at an alarming incident rate of 50%.
Despite the failure of AI to deliver expected results, the Fivetran report shows that nearly nine in ten organizations continue to use AI/ML methodologies to build models for autonomous decision-making. Ninety-seven percent are planning to continue or start investing in GenAI in the next 1-2 years.
“The rapid uptake of generative AI reflects widespread optimism and confidence within organizations, but under the surface, basic data issues are still prevalent, which are holding organizations back from realizing their full potential,” said Taylor Brown, co-founder and COO at Fivetran.
Taylor further added that “Organizations need to strengthen their data integration and governance foundations to create more reliable AI outputs and mitigate financial risk.”
The report also highlights the dissonance between various job roles. Technical executives, who build and train AI models, are less convinced about their organizations’ AI maturity. Senior executives feel the lack of AI skills is a greater obstacle to AI adoption, while decision-makers in more junior roles believe outdated IT infrastructure is the top concern.
A key reason for underperforming AI programs is the quality of data in terms of its accessibility, reliability, and accuracy. The growing number of GenAI use cases has further exacerbated the issue with data quality.
With organizations looking to increase AI infrastructure in the next few years they must find solutions to overcome these challenges. Having solid data governance foundations and following good data practices could be a good starting point to lay the groundwork for successful AI deployment.
Related Items
Why Digital Transformations Failed and AI Implementations Are Likely To
Altair Survey Unpacks the Friction Surrounding AI and Data Projects
The Three Approaches to AI Implementation
#AI/ML/DL #AI #genAI #IT [Source: EnterpriseAI]