Analytics technology revolutionized how businesses mine data to unlock hidden value, and now Generative AI (GenAI) is further disrupting predictive and prescriptive analytics. Companies are experimenting to various degrees with dozens of GenAI platforms. The most widely known is ChatGPT, an internet-based AI model that mines unstructured data to mimic natural communication — the basic function of all such platforms.
However, professionals are still exploring GenAI’s varied applications in data analytics. On the one hand, proponents say integrating its ability to sort through unstructured data with analytic technology’s proven capacity to identify patterns, trends and anomalies in organized data is paying dividends.
On the other, research suggests there is significant hesitation in the C-suite about cybersecurity and data quality, which can lead to flawed or illegal decision-making. “To realistically gauge [GenAI’s] plausible impact, we must take a closer look at the mechanisms that translate technology to broad productivity growth,” the World Economic Forum suggests.
What Is GenAI?
Up to 90% of existing data is disorganized, according to TechReport, and almost all businesses surveyed say that presents a nearly insurmountable challenge to extracting meaningful, high-value insights from it. Texts, images, voice communication, printed documents, videos, emails and instant messages are sources of unstructured data that contain trillions of dollars of unrealized information. The contents are difficult to organize, analyze and store in structured databases that computer and analytics software require.
GenAI changes things, however. “The emergence of AI and machine learning has paved the way for novel software tools that efficiently navigate extensive volumes of unstructured data, unveiling valuable and actionable business insights,” TechReport predicts.
How Can GenAI Streamline Predictive and Prescriptive Analytics Tools?
As students learn in programs like Radford University’s online Master of Business Administration (MBA) with a Concentration in Business Analytics, businesses use data analytics to answer four types of questions:
- Descriptive: What happened?
- Diagnostic: Why did that happen?
- Predictive: What might happen next?
- Prescriptive: What are the options for addressing predictive possibilities?
GenAI’s primary value is its ability to organize unstructured data to support predictive and prescriptive analytics. Data used in descriptive and diagnostic analytics is historical and already structured.
Forbes cites GitHub, a cloud-based software and version-control platform, as a successful example of GenAI supporting predictive analytics. The magazine reports that developers who use GitHub’s AI-powered platform report significant gains when creating code and datasets to train AI and machine learning models. A survey found that 88% of respondents indicated increased productivity, and 96% said GenAI expedited repetitive tasks.
Logistics and supply chain operators are integrating GenAI and prescriptive analytics to adjust shipping, warehousing and demand forecasts in near-real-time with the volatility and unpredictability of global markets, according to Logility.
Shifting to AI-powered supply-chain planning enables shippers to anticipate demand fluctuations and “fine-tune … shipping volumes, vessel capacities, and port limits … slashing transit durations by 20% and decreasing fuel usage by 15%,” according to the supply-chain optimization vendor.
Challenges of Adopting GenAI in Business Analytics
Despite those promising results, the Boston Consulting Group (BCG) 2023 Digital Acceleration Index found most of the 2,000 global executives surveyed do not fully understand GenAI and discourage its use across their enterprises. Among their top concerns were data security and privacy, bias and ethics and intellectual property risk. Accuracy is another issue, as there are no processes currently to backtrack the GenAI-produced content to verify the quality of the data used in large language model (LLM) platforms such as ChatGPT.
Hackers have also discovered ways to attack LLMs by methods such as prompt hacking to produce false or misleading information. When coders use such information to write programs, they inadvertently create potential cyberattack vulnerabilities. “The impacts of large language models and AI on cybersecurity range from the good to the bad to the ugly,” InfoWorld warns. “Any tool that can be put to good use can also be put to nefarious use.”
Nevertheless, the BCG survey found that business leaders envision significant returns on their investments in GenAI. If these returns occur, businesses will “have even more incentive to tackle their challenges and concerns.”
Students in the online MBA with a Concentration in Business Analytics program from Radford University are equipped to adapt to GenAI developments and trends and leverage such resources to benefit their organizations and create career opportunities.