It seems that most conversations involving technological innovation are now dominated by these two words: artificial intelligence. Advocates maintain that AI’s ability to sort through petabytes, exabytes, zettabytes, etc. (well, you get the message) of data in a nano-fraction of the time it would take a human opens the door to huge amounts of potential solutions or strategies for businesses. For fossil fuel companies that are steering their way through the transitional energy forest, AI promises to provide some predictability, and maybe even control, to come out the other side of the forest in a healthy state.
Ana Domingues is EY’s AI Lead for Global Energy and Resources and works with clients around the world to develop their AI strategies. At the recent Energy Disrupters Conference held in Calgary, Ana discussed this pressing issue as a keynote presenter with fellow panel participants but also took few minutes to talk about AI with Business Edge’s Ernest Granson.
BE: Ana, really appreciate you sitting down with me today. You are very passionate about the potential for AI to not just contribute to better energy management, but to actually lead the way for a successful transition, and you've been delivering that message around the world as recently as this past summer at the World Energy Congress. But what is it about this technology that has you feeling so strongly about it?
AD: Well, it's about the technology’s unique ability to analyze through complexity, the ability to understand patterns and trends and go through millions of potential possibilities for a solution, then identify the one that fits best a specific problem, doing it in minutes and not in hours, weeks or years, and doing it with the accuracy that is required to provide confidence and minimize the risk of the decisions we are making.
BE: It's my understanding that all the biggest oil companies are already using AI in some form or another: ExxonMobil Shell, BP, Chevron, Total Energies, etc. What are the practical ways in which they're utilizing the technology, and how is that contributing to the actual transition?
AD: All those big companies you've mentioned, and many others, are big champions of AI. They have been exploring AI for many years already. Aramco, for example, announced about six months ago in a press release, that their AI solutions have delivered them $500 million in savings. AI was used to increase productivity, reduce maintenance costs through predictive maintenance, and also to increase safety. As well, the implementation of AI on the upstream component helps to increase energy efficient exploration which contributes to a decarbonizing effect.
In some cases, 50% of refinery OPEX are from the energy consumption. Imagine if not only savings can be gained from being more efficient, but also what you can do for the environment by consuming fewer commodities such as electricity and water. AI can be used to automate the activation of measures that create those savings. These are just some examples
There are also other ways that AI can benefit decarbonization, specifically, support for carbon capture. There are multiple studies from the International Energy Agency that points to the fact that we will consume oil and gas for a long time. So, the route to energy transition is to decarbonize, to make greener those oil products for that you can use carbon capture. Carbon capture is a very expensive technology and it is still embryonic. AI can help companies that are already developing this technology in multiple ways. It can also help to find and identify reliable repositories of captured carbon. For instance, AI can help with the analysis of potential cracks in the repositories in order to confirm the trust that you must have in such storage systems.
These are just some practical examples where AI can help with the current extraction of fossil fuels and in the transition of bringing in the renewables of sustainable energy. They are the two key aspects in which AI can contribute: to make the consumption of energy more efficient through the whole value chain of the oil and gas, such as preventing spillovers, reducing flare events and venting, all of which are contributors to emissions; and to help advance carbon capture to make it more profitable and less risky.
But to speak to renewables as you mentioned, many of the oil and gas players have started building a renewables portfolio. AI has multiple uses in helping to make renewables businesses more profitable and more acceptable from a social and environment perspective.
BE: What about AI’s use for mid- and small-sized energy companies, whether they are in the sustainable energy sector or the legacy fossil fuel sector. They may not have those financial resources to use this really advanced type of AI,
AD: That's a very good point. Here, at EY, we have many discussions about the practicability of AI. Our recommendation is to look at what these giant oil and gas companies have been doing in that area and how it can be leveraged as a proven success. The benefits are clearer and so there is less risk for the smaller companies to follow. This is important because we need to show that AI and other technologies can actually deliver on the returns for the smaller companies. Generative AI models have already been trained, leveraging the large language models, so it's a faster route to value and impact.
BE: Well, let's talk about the elephant in the room, which is energy consumption and the many questions being asked about it. Will the advantages offset the disadvantage of that huge energy consumption?
AD: I am usually optimistic by nature. So, when I think about the level of consumption, it’s positive that the companies pushing to use AI are giants in the industry. They have huge investment and research capabilities and they also have very ambitious sustainability goals. In terms of energy consumption, it’s fortunate that AI is mostly in their hands, because they will have the capabilities to advance research sufficiently to start reducing the consumption of energy. I would say that this large energy demand that is coming will be flattened and reduced as more efficient models are developed. There are multiple research efforts already taking place.
Is the current energy consumption worth it? I have a strong belief that there is no energy transition without AI. The level of complexity towards which we are heading is extraordinary and the fact that we must make decisions in minutes, so much more quicker than we have been doing, convinces me that only AI can help us with it. We humans, need to have our capabilities augmented to be able to go through the very complex energy system towards which we are heading.
And it's not just operations. It's also about making all the elements profitable fast enough so that we can deliver on the energy transition. The second thing is, and here I'm going to steal some insights from Microsoft. Generative AI actually gives the dream of every single person on the planet to have a doctor and a teacher, because with Gen AI, you can start having access to medical help and education without having that person in front of you. That kind of help is only feasible, for 7 billion in the world if you use AI. I think that's good dream, good enough to pursue and invest in making AI more efficient. That’s not a concept I expected to be brought up, but is something important to be thinking about. It’s a concept of optimism.
BE: Currently, we're really only seeing what a fourth generation of AI can do. Considering how quickly these new generations are being developed, do you think, or should we be expecting an even greater leap in capabilities as these generations become available? If so, what possibly more can the technology offer?
AD: Wow, that's a very difficult question. In terms of maturing and evolution of AI, it's going to be even more exponential. Humans are like this, right? We like to invent things, and we are just scratching the tip of the iceberg on what AI can do. Many people say that AI will eventually be able to completely replace humans. Maybe, I don't know. Maybe it will take 100 years to get there.
I believe that AI is an extraordinary tool to augment people's capabilities. You will always need humans in the loop, but AI will start doing more and more things that humans are doing now – things that are mundane, repetitive, boring, and make your brain stiff, right? Having said this, the transition, as in any transition, will require people to invest energy, time, bandwidth and the will to change. That is the complexity of the journey. It's in the people who will be changing this technology. It’s a human-centric transition a human-centric journey, because people will need to change dramatically, the way they work, the way they do their activities in professional and personal lives and, unfortunately, like with many other technologies, those that don't progress along with it, will remain on the outside.
BE: That's a really powerful perspective, which I think, around which we all have to wrap our heads. That brings me up to this next question. In your experience, have you run into what you would call cultural resistance to the use of the technology and the way it's applying to transition.
AD: Oh yes. Every day. It's part of human nature that we embrace curiosity, but we are made of habits, right? Changing habits is hard, and that's why, at EY, we talk so much about change management. It's not that people do not want to change. It’s that it requires a personal investment and also time required to think about how you're going to change, how you're going to do things differently. There is always a temptation for the majority of the population to try to do the things they have always done. So that means multiple examples of resilience.
It varies a lot with the companies I with whom I consult. An example is the CIO of a major electrical company, with who I discussed on where to start when investing in Gen AI. He said, ‘Let's not start with the engineers. Engineers are going to doubt everything we give them.’ That was an interesting claim, and it is true. If you give a predictive maintenance model to an engineer and there is no trust, the engineer will continue to do the things exactly the same way it has been done the last 20 years. Forget the investments you've done on the model. You know, there is no value realized until it gets adopted in another company or a completely different department. Here it was the CIO, who was saying there will be a big challenge in introducing generative AI because not even the CIO believes in it.
My comment was that, in just a couple of years, the CIO will see himself only with colleagues that are older than 40 years of age because no one younger will want to develop software and be part of an IT group if generative AI has never been accepted in that department. Every company needs to find the right balance of change. If you don’t have deep pockets, you don’t want to move too fast because you can really fail quickly. But you also don’t want to go too slowly because you may miss the train. That will be a disadvantage if you are trying to attract the talent such as engineers, field technicians and data scientists. Many industries are struggling to attract that talent.
BE: We're still dealing with the human nature aspect. I would imagine that's the hardest part, not necessarily developing the technology.
AD: Well, developing the technology, is not easy either, right? But there is no point in investing so much effort in the technology if the human part is not at the centre and at the beginning of thinking about what to do with AI.
BE: Anna, thank you so much for providing this insight. It's such a rapidly advancing technology, we need all that information we can get. Thank you very much for being here with us today.