Witthaya Prasongsin/Getty Photographs
Information is commonly referred to as the brand new oil due to its worth to trendy enterprise. Within the case of ExxonMobil, senior IT government Andrew Curry says knowledge is not simply the brand new oil, however the gas for a complete vary of AI-led initiatives throughout the corporate.
As supervisor of the oil and fuel big’s Central Information Workplace, Curry is chargeable for executing enterprise-wide knowledge ideas — and the speedy development of curiosity in rising applied sciences, similar to AI and machine studying (ML), in the course of the previous 12 months hasn’t escaped his consideration.
Additionally: If AI is the way forward for your online business, ought to the CIO be the one in management?
“Everyone’s speaking about AI and ML, however you may’t achieve success in that space with out having high-quality knowledge,” he says, talking with ZDNET in an interview on the current Snowflake Summit 2023 in Las Vegas. “The extra alternatives that we’re seeing in AI, the extra that is placing stress on us to ask, ‘Do we’ve got the standard knowledge to achieve success on this space?'”
Curry has refined his reply to this query whereas working for ExxonMobil since 1999. After filling a plethora of roles, and serving to the corporate set up its Central Information Workplace, he is realized that one aspect is essential to success in rising know-how: knowledge technique.
Companies will not be capable to benefit from AI and ML except they first guarantee they’ve a powerful, safe knowledge technique, which is one thing Curry has established at ExxonMobil, each throughout his earlier roles and since shifting to his present place because the agency’s knowledge chief in Might 2023.
“We perceive the standard of our knowledge,” he says. “We all know which knowledge is prepared for superior AI and ML capabilities. We additionally know, and that is equally vital, which knowledge is not fairly there as properly when it comes to high quality.”
Additionally: AI gold rush makes fundamental knowledge safety hygiene crucial
As supervisor of the Central Information Workplace, Curry is establishing a long-term knowledge technique that defines the applied sciences, processes, skillsets, and guidelines to assist the group handle its info belongings efficiently in an age of AI.
As a part of its knowledge technique, ExxonMobil has a know-how ecosystem that makes use of Snowflake’s cloud-based platform. Curry says Snowflake has given the enterprise a consolidated knowledge basis for the primary time.
Extra usually, his staff helps guarantee any knowledge initiatives are proper for the enterprise. A giant a part of that effort focuses on the areas the place ExxonMobil can differentiate itself by way of knowledge — and the insurance policies that may permit folks to take advantage of rising know-how in a secure and safe method.
“It is advisable know the place you may leverage this know-how and know the place you are not prepared but,” he says. “We’re actually positioning ourselves in the best areas.”
Additionally: Generative AI is the whole lot, all over the place, unexpectedly
To the top, Curry says his staff is evaluating giant language fashions (LLMs) and generative AI capabilities as a part of its long-term knowledge technique.
Proper now, there’s not a lot to speak about when it comes to public developments — as soon as once more, the main target is on guaranteeing the agency’s knowledge high quality is excessive earlier than plowing into initiatives.
“I will say there’s nothing in manufacturing but for us,” he says. “However we proceed to place that know-how and to know the place the enterprise alternatives are, and we’re ensuring the info is prepared for that work.”
Like so many different executives at giant enterprises, Curry is taking a cautious method to the usage of LLMs and generative AIs as a part of his firm’s knowledge technique.
Additionally: Open supply is definitely the cradle of synthetic intelligence. This is why
“Generative AI and LLMs is an space the place we have put a lockdown on it,” he says. “At ExxonMobil, you may’t simply exit and use this know-how.”
Different analysis suggests a cautious method is much from uncommon. In reality, many corporations aren’t even able to discover rising know-how.
Whereas 98% of executives polled in a current survey by Workday imagine there are doubtlessly large advantages from deploying AI and ML, nearly half (49%) report their group is unprepared to take advantage of these rewards because of a scarcity of instruments, expertise, and information.
For Curry, these are vital hurdles that should be overcome earlier than a blue-chip firm like ExxonMobil begins experimenting with generative AI and LLMs in a manufacturing surroundings.
Additionally: AI goals to foretell and repair developer coding errors earlier than catastrophe strikes
“I feel there’s lots of points about securing your knowledge and serious about what turns into public info if you leverage a few of these public instruments, for instance,” he says.
“So, we’re taking a cautious method for broad use inside the firm, however at similar time, there are energetic plans alongside the way in which in a extra managed surroundings.”
Samantha Searle, director analyst at Gartner, agrees that almost all blue-chip companies are taking a cautious method to generative AI.
Like Curry, she says getting your knowledge technique proper is a vital first step for organizations earlier than they begin pushing knowledge into LLMs.
“Completely, as a result of folks aren’t going to love it if you’re recommending the improper issues to them,” she mentioned to ZDNET in a video interview. “That is going to have the antagonistic impact when it comes to buyer retention.”
Additionally: How trusted generative AI can enhance the linked buyer expertise
Searle additionally factors to the significance of information high quality. Proper now, digital leaders must be concentrating on how their info belongings are collected, saved, and exploited.
“It is crucial to guarantee that what the AI fashions predict is correct. We nonetheless want people to confirm the outcomes are proper — we all know with generative AI that these instruments can hallucinate,” she says. “So, you continue to want verification steps. And ensuring you’ve got bought optimum knowledge high quality is a key step to mitigating these dangers.”
Again at ExxonMobil, Curry says it is vital to acknowledge that, whereas his enterprise is cautious about generative AI, it has a protracted historical past of using different areas of ML and AI.
He expects the corporate to make use of rising applied sciences in a number of key areas going ahead.
First, finance and buying and selling: ExxonMobil has a big money circulation, and AI and ML might assist refine the agency’s processes.
Additionally: Generative AI tops Gartner’s high 25 rising applied sciences for 2023
“Are we uncovered in sure areas? Do we’ve got an excessive amount of money in sure areas? Ought to we be investing in different areas? We predict we will make enhancements on these margins with our money if we will leverage our ML capabilities.”
Second, provide chain: “With the ability to perceive when issues are being impacted is essential, and having the ability to react in a well timed method means quite a bit to the enterprise, and so that could be a key space that we’re investing in.”
Lastly, Curry factors to the potential use of rising know-how on the subsurface stage to assist interpret traits in seismic surveys.
“That is an space we’re working to progress,” he says. “It is most likely one space that is going to have an extended lead time for us as we work to know it.”