DTECH 2024: IBM is at the Forefront of Practical AI in Energy with LLMs for Data Analysis and Decision-Making

Since ChatGPT's launch in 2022, the AI landscape has been in a state of rapid evolution. Google's Gemini and OpenAI's text-to-video generator Sora are just a few examples of how the AI space is heating up. Nowadays, businesses are transitioning from experimental AI endeavors to investments in technologies that deliver immediate operational and business benefits. One of the key trends is the adoption of large language models (LLMs) to improve data analysis and decision-making processes.

These LLMs, which are trained on vast datasets, are now being fine-tuned to handle specific time series and operational data, ensuring that sensitive information remains secure within organizational boundaries. This approach not only augments the capabilities of existing models but also addresses concerns about data privacy and security. As industries continue to recognize the potential of LLMs, the emphasis is on harnessing these advanced technologies to drive innovation and efficiency. To delve deeper into this topic, MarketScale spoke to Casey Werth, Global Industry GM for the Public Sector at IBM Technology, on the show floor of DISTRIBUTECH 2024, the four-day energy transmission and distribution exhibit held in Orlando.

Werth's Thoughts

Transitioning from Experiments to Investments

"I think that everybody's looking for, what are the good ideas of where to start where there's going to be quick creation of business or operational value. I think that we've sort of, in the last six-seven months, moved from science experiments or trying things out and now people are really looking to invest."

I think that we've sort of, in the last six-seven months, moved from science experiments or trying things out and now people are really looking to invest.
— Casey Werth, Global Industry GM for the Public Sector at IBM Technology

Leveraging and Enhancing Large Language Models

"We're looking actually at now leveraging large language models to leverage the strength of large models that have already been trained but to look at time series or other operational data sets, allowing data to go out of the walled garden if you will or the firewall and so absolutely there's a focus now on bringing in the best capabilities but then training them and making those models more valuable within the organization and making sure that that data doesn't get out. Of course, it increases the value of the model so that's absolutely a focus of most clients we speak with."

It increases the value of the model so that's absolutely a focus of most clients we speak with.
— Casey Werth, Global Industry GM for the Public Sector at IBM Technology