Artificial intelligence is not new. The mathematical and theoretical concepts behind today’s AI algorithms date back to the mid-20th century, but in the last decade, big data and new technologies have advanced rapidly to start delivering on AI’s promise.
THE GAME-CHANGING NATURE OF GENERATIVE AI
The dramatic growth of generative AI tools over the last year or so has further demonstrated that AI is not a passing fad. In 2022, data from McKinsey’s global survey on AI showed that AI adoption in organizations had more than doubled since 2017 but had plateaued between 50% and 60%.
In McKinsey’s 2023 survey, published less than a year after ChatGPT launched to the public, 79% of all respondents across regions, industries, and seniority levels said they’d had some exposure to GenAI. A third of respondents said their organizations are regularly using it in at least one business function, and 40% reported that their organizations will increase their overall AI investment because of GenAI advances. AI adoption is gaining momentum in all sectors, including the energy industry.
“When a software gets mass adoption, it’s because everybody finds it easy to use,” said Jay Bhatty, CEO and founder of NatGasHub.com, a company that provides automation software for the oil and gas industry. “Generative AI is a democratized software, which is why it’s a game changer. I don’t even use Google search anymore; I use ChatGPT. It gives you much more relevant answers on a plate instead of having to read 10 different links. AI is the future, and it’s already here. It’s been in use for a long time, but the difference now is that use is accelerating. It’s just a matter of people upskilling and learning how to use it.”
AI APPLICATIONS IN THE ENERGY SECTOR
A major challenge for companies in the energy industry is aggregating, processing, and analyzing large amounts of data from numerous sources. Without AI tools, company employees are responsible for doing this monotonous and time-consuming work manually.
“People in our industry have many vast repositories of data, some that are available on the public internet and others that are within their own company intranet,” said Bhatty. “On the internet, for example, we deal in data that is thrown across 500 different natural gas pipeline websites. People have to manually scrape this data from all these websites every day—and that involves a lot of manual, repetitive tasks. But now, our AI software automatically scans these pipeline websites every day and tells our analysts if something has changed. If there is a change, analysts can go directly to the source without having to look at all 500 websites.”
NatGasHub.com uses a language learning model built on OpenAI’s API and customized with proprietary code. The LLM is trained to consume and understand data the way a human would, only much more quickly, and then perform tasks on it.
“It’s important to understand that you don’t have to create a language learning model from scratch,” said Bhatty. “You use what’s already available, built, and proven, and then you refine it. Using AI has shaved off a lot of time for employees who can now focus on more mentally challenging tasks—and if you add up these time savings and multiply them across several employees, you see huge time savings and productivity gains.”
STRATEGIES FOR SUCCESS
Bhatty recommended these three strategies for leaders in any industry interested in adopting AI tools in their organizations.
1. ADDRESS PRIVACY CONCERNS
Privacy is a legitimate concern for companies using generative AI. With free GenAI models, such as ChatGPT, queries become publicly available data.
“To protect sensitive or proprietary information, consider getting a paid subscription to one of the products that allows you to ring-fence your data,” said Bhatty. “Then you can have the model analyze your data privately and not release it to the public.”
2. START SMALL
Implementing new AI tools can seem intimidating at first. By adopting a startup mindset and aiming for a minimum viable product, leaders can avoid scope creep and establish a track record of success.
“I tell clients to start by addressing a small, manual pain point in their companies,” said Bhatty. “Don’t keep developing forever. Get feedback, improve, and iterate. Prove that AI is helpful with small problems, then tackle bigger problems.”
3. GET YOUR TEAM INVOLVED
Some companies are hesitant to encourage employees to use new technologies because they are unsure of their value. But many employees are tech-savvy and eager to explore new tools, and embracing this curiosity can foster innovation and contribute to positive business outcomes.
“Make AI technology available to your employees, even if you’re just testing a free version,” said Bhatty. “Let them try new things, fail, learn, and try again. Encourage them to think of AI as something that can augment their performance and skill set. I guarantee that if you’re not allowing them to use it at work, they’re using it at home anyway. Technology is like time; it goes in one direction, and either you’re on the right side of it or you’re not.”