Navigating the Landscape of Generative AI: Opportunities and Risks – A Discussion with Rachel Stuve
Generative AI has been the talk of the tech world for months, but how are companies using this technology in the real world? Our Director of Mid-Market, East, Steve Wright, sat down with Rachel Stuve, Senior Director of Data Science & AI at Elevance, to speak about Generative AI's business potential.
AI in Simple Terms
Rachel sees AI as the overarching umbrella of several technologies, including large language models (LLMs), computer vision, and machine learning. Each of these technologies take massive amounts of data and finds patterns in them.
“The fundamental use of data is to make business decisions, and how you harness this data will change as the industry changes,” shares Rachel. Right now, more businesses are attempting to harness this data with AI, whether it’s finding patterns in numbers for financial purposes or using generative AI to find patterns in text.
Challenges of Generative AI and How Companies Plan to Overcome Them
AI relies on model training to know what to look for, and training models requires massive amounts of data. There’s certainly no shortage of data out there, but the need to access such large amounts of information has drawn several concerns.
The accuracy of data being released to these models is questionable, and it has real business implications if left unchecked. For instance, Lyft recently released its quarterly earnings but added an extra zero. The company quickly retracted the report, but damage could be dealt quickly if this data were absorbed by AI models. This is just one example of how LLMs rely on accurate information— If it ingests bad data, there is a greater chance of inaccurate outputs in the future. “It only takes one or two really bad offenses to crash the house of cards.” To battle these and other challenges, many companies that started with public LLMs like OpenAI are starting to build internal versions of these models to harness their own intellectual property and data. Rachel predicts these companies will be the most successful in deploying AI because they control their own information in their own ecosystem while selectively using public data. “It’s like a two-way mirror, where there’s information coming in but not out.”
Predicting the Future of Generative AI
Companies are still rushing to adopt AI, mostly out of a perception of need. “There’s FOMO with competitors and lots of hype still, plus there’s a financial upside for publicly traded companies that want to see their stock prices increase by saying they’re using AI.” While there are still nuances to consider, like bias and personalization, Rachel believes generative AI holds major value. “I’ve seen it deployed to check the status of a claim, summarize content, and speed up the creative process without replacing creative people,” she says.
Rachel predicts the difference between success and failure when using AI is to be intentional about its benefits and use cases, not just replace headcount. Getting around the challenges and helping people find ways to use AI to its full potential will be key to AI’s staying power.
Connect with Rachel on LinkedIn.
Share this article
Let's get started.
"*" indicates required fields