Unveiling the Key to AI: Discover How Data Completes the Puzzle and Bridge the Gap

 The challenges holding back progress in artificial intelligence (AI) are not just about a shortage of skilled professionals. According to a recent study by IBM, there's another big hurdle: data complexity. The study shows that the top obstacles to AI success are limited AI skills and expertise (cited by 33% of respondents) and too much data complexity (25%).

Surprisingly, a majority of companies (58%) haven't fully embraced AI yet, according to a survey of 8,584 IT professionals. Among these companies, the main roadblocks to adopting AI include concerns about data privacy (57%) and issues related to trust and transparency (43%).

For those companies already using AI, the challenges often revolve around data. Some are taking steps to ensure trustworthy AI by tracking data provenance (37%) and reducing bias (27%). Additionally, about a quarter of companies (24%) are focused on developing their business analytics or intelligence capabilities, which rely on having consistent, high-quality data.

However, some industry leaders are warning that organizational data might not be ready to support the growing ambitions of AI. Matt Labovich, US data, analytics, and AI leader for PwC, emphasizes the need for CIOs and technology leaders to adapt their data strategies to integrate AI effectively.

Shipra Sharma, head of AI and analytics at Bristlecone, highlights the importance of addressing data security, AI decision-making ethics, and AI literacy. She suggests that actively engaging with the technology and implementing appropriate safeguards will help organizations harness the benefits of generative AI for data management while minimizing risks.

To make progress in AI, Labovich suggests that companies need to strike a balance and recognize the significant role of unstructured data in advancing generative AI. Sharma agrees, noting that organizations don't always need to use generative AI on top of structured data to solve complex problems—simple applications can sometimes be the most efficient.

The variety of data required for AI can be challenging, with data at the edge becoming a significant source for large language models and repositories. Bruce Kornfeld, chief marketing and product officer at StorMagic, emphasizes the urgency for companies to find cost-effective approaches to filter out unnecessary information and make room for essential data.

Osmar Olivo, vice president of product management at Inrupt, highlights the dilemma many organizations face in choosing between leveraging AI for a competitive advantage and protecting sensitive data. He anticipates innovative data management and privacy solutions emerging in 2024, particularly focusing on safeguarding data used by AI models.

Rakesh Jayaprakash, chief analytics evangelist with ManageEngine, stresses the importance of a data-first approach and a centralized data repository for successful AI adoption. He suggests capturing every organizational event and process, using machine-learning algorithms to discern valuable patterns.

While the future promises features with generative AI, businesses should exercise caution when investing significant resources. Jayaprakash advises organizations to make no-regret moves, like streamlining operations and implementing short-term improvements, while larger data and technology initiatives are in progress. This, he believes, can yield benefits such as enhanced productivity and cost savings.