← CERAWeek Podcast with Atul Arya
CERAWeek Podcast with Atul Arya
David Rabley, Accenture, on Why AI Is a Business Strategy, Not an IT Project
00:19:57
Loading summary
Transcript30 lines
- [00:00]
A
Foreign.
- [00:07]
B
Welcome to the sarawik Podcast, where the world's energy leaders and innovators share insights on the future of energy, technology and climate. I am Atul Arya, Chief Energy strategist at S and P Global. In each episode, we dive into the critical issues and bold ideas shaping our energy future.
- [00:27]
A
So let's get started. Hello everyone. Welcome to sarawik Podcast and today we are actually coming to you from Sarawik itself. It's day one of Sarawik 2026 in Houston, Texas. And joining me today is David Rabley, Global energy strategy lead at Accenture. David, welcome to Sarawik and to this podcast.
- [00:55]
C
Ato. Thank you. And to everyone out there listening, great to be with you.
- [00:58]
A
We're going to talk a little bit about, of course, AI, which is the big topic, but with a twist. How are companies getting ready for AI creating an AI ready organization? Already on day one, I've been hearing everything about how great companies are doing and how much value they are generating, but I'm sure we need to clarify what they are doing. So how are companies doing today? Give us kind of an overview of
- [01:22]
C
what you're seeing in the marketplace here on day one. I've got to say, probably the number one topic that I'm hearing about in many of the sessions today is gas, let's be honest. But alongside of that, the conversation really quickly pivots to what is happening in terms of AI, where it can be a solution. And I've been coming to Sarah week for a number of years and I have to say that each time I step forth here, I'm really inspired by the examples that the leaders are sharing on the stage around where they're using AI. If you stepped back maybe a couple of years, I felt that we would talk about it as a thing for the future. We would hear examples in the back office, maybe we would hear examples around legal or other support functions. But today, I think without question, AI is now on the lips of every leader and they're talking about how it's changing the business that they are running. It's talking about how it's impacting cost, productivity recovery, performance across oil and gas. I was in a session with you this morning and I think we realized that it was going to be super important that going forward, energy companies are driving production and they're driving efficiency. And I have to say I don't really think that we've got a better answer out there than taking that next level of performance through AI with AI and going forward.
- [02:43]
A
Yeah, good progress so far. So how are companies setting their AI strategy, how advanced they are thinking on building an AI strategy, linking it to their business strategy?
- [02:54]
C
I love your point on linkage. And our perspective is that an AI strategy is a business strategy and it has to start with value. Many times the question about AI comes from many sources within a company. But where it has proven most effective is where it isn't considered just to be a technology story or it's a question of an IT topic. It's actually a question of business value. And that makes it perhaps unique in terms of a general technology that is really going to make a difference throughout the organization. And if I can say a little bit more about that, I think those companies that are now showing genuine outcomes and they are really in a minority, it's still a minority. It's around 10% of enterprises are showing results grade performance from AI. But where they're doing it is when they're embracing this not just as a technology project or a data project, but they're seeing this as part of what they need to drive holistically for their business. So if I could say a little bit more about that at all, what I think is really the difference then is is that those companies have the sponsorship at the very top. Those companies are challenging and rewriting the rules for who is driving those gains and owning that delivery. And we're having to see a change in organizations from that. And I was thinking a little bit as we're walking around zero week today, around the sort of examples, back to your first question, as it were, that we're hearing of. And in some ways it's quite obvious to me that AI can do an awful lot of things better than humans. And what I would like you to start thinking about a little bit here is that when an oil and gas company is embracing AI, it's looking at its strategy. For AI really, it's got a choice. It can do things that it's already doing, but it can do them better. By that, it could have a faster outcome. It could have less cost associated with it. It could have fewer resources tied up. It could also do something better. So it could optimize across wider range of challenges. It could bring more data into the mix. It can actually put together parts of their business that were separately existing or siloed before. But importantly, AI can do things that were not possible to do in the past. The power of AI isn't just going to be doing what we're doing today for less people or less cost. It's actually going to be when you start putting together processes and groups from across companies that never connected before. There are some great examples I heard on the stage today and that we work with where oil and gas companies are now saying, wait a minute, can I go back to some of my historic logs? Can I re look at some of my seismic. Can I re challenge myself to see did I leave resources in the ground? If I could benefit from putting today's technology against yesterday's information, would I come to the same conclusion? And so I think a lot of what we consider to be the domain of AI today is what does it mean for productivity and resources and cost. But I think increasingly I'm really excited about AI doing something that we didn't do in the past or doing something in a way where the outcome is better than we could have imagined.
- [05:57]
A
So very holistic way of thinking, very differently. So tell us, what does an AI ready organization look like to you and how different will it be to from organizations today?
- [06:08]
C
It's an amazing question. If I knew completely the answer at all, it wouldn't still be up for debate. But already it's becoming clear that an AI organization could and will look very different to any organization that we've seen today. Now, what does that mean at a certain level, companies come to us and they're asking, do I need the same workforce? How many people do I need in the future and do I need the same skills? And so there's category one is how does the workforce change? And you hear a lot about this. Are there jobs for today's graduates in the future? Are we going to have the right leaders coming through? How can we get AI fluency through our organizations? What do people need to be doing differently? How can they add value to this? And we're finding increasingly that, yes, there will be a shift. For example, some work we did recently showed that for a super major you could have 25% fewer roles. But the story doesn't end there because within that pyramid, it moves to being a little bit more senior, it moves to having more judgmental roles. It moves towards having more human decision making, more expertise. And so the first thing is the workforce will be different. The second thing is that the work itself is going to change. So when you consider today, an organization like an oil and gas company is a tremendously complicated set of tasks. And I don't think any two players or any two people attending ciraweek this week have exactly the same mix of assets or participation across the value chain. But interestingly, they're all composed of a set of processes. And as you deconstruct any organization and you ask why do you do your work this way? The answer you often get is, this is what worked for us because of the systems and the technologies that we grew up with, or we're using or we've adapted to. In the last couple of years and last couple of months even, I think we've seen a real reset in terms of the capabilities and how work can be done. And so the second thing that's going to be different about organizations is they're going to look at those processes. And our advice is not to take too many, maybe 6 to 8, that drive 50 to 80% of the value of the company and ask yourselves the question, could I do this work differently? The third thing is structure. Organizations can look radically different. I always think about why does an oil and gas company look the way it does? And my theory was that it comes from the domain expertise of what people studied in university and the disciplines associated with that. I don't know if there is in fact another industry which has remained so tightly focused on the geology and the geophysics and the geosciences as the oil and gas industry. But the thing is, these domains are now changing. The reasons why you needed to structure, whether it was by geography or it's by process, or it's by value chain, have gone away in many extents and knowledge has become much more broadly accessible. Institutional capabilities are completely different. And so there really is no reason to be as siloed tomorrow as we are today. And so we're anticipating a much flatter, less level of organization. And I think certain constructs that were necessary because of the disciplines, the changes of knowledge across the organization, they're going to be challenged, they're going to go away, and we're going to see a much more unified organization structure going forward. I would come back to the example where you imagine a couple of hundred years ago, in fact, when steam power was in its early days and factories were essentially vertical, they were vertical because once you generated the power from a steam engine or even a water wheel, you had a central drive shaft drove power through that factory. And the most efficient way to do that was to have a vertical factory with the roof stacking on top of each other. With the advent of electricity, that changed. You didn't need to have buildings that were small in footprint, but stacked. You could have what we would recognize today as modern factories, so large single story buildings where the power is distributed and you can organize your production in a much more efficient way. Think about the analogy here. With an organization, we've grown up in a world where we needed to have that vertical control and that governance and that drive shaft coming down through the organization. We're now going to be in a position where AI can provide that knowledge at the point at where it's required. Each individual's capacity to take that knowledge is so different and the spans will be very much changed. So that's really the third thing that the structure will change. And then I think beyond that, you've got to really ask the question of how decisions are going to be made in an organization. And we're going to move from today's world of experience and hierarchy to what Accenture refers to as a digital brain. What that means is that companies are going to start really thinking about how can they make complex trade offs and decisions across their organization that really take the context and the opportunity from one part of the world and deploy it to another. So think for a second around our industry. We're continuously disrupted, but we often still have individual assets that are having to live with their annual plan, with their strategy, with their metrics. And that means sometimes we're optimizing for that asset, we're not optimizing for the system. But it is very complicated to build a system of trust and of governance by which you could say, hey, wait a minute, we're going to run this asset differently because you're part of a system, but I can also reward you for that and I can incentivize you for that and I can give you the right levels of decision making that come alongside operating in this different way. And so those organizations which I think are going to really emerge in the future are the ones where the digital brain is smart enough to make those trade offs. And the other way to frame it, if you prefer more tech language for this, is that is your agentic stack. That is the equivalent of taking your data, your ontology, your semantic layer and your agentic architecture, building that through and being able to use it to make decisions. I don't mind which way we frame it, but those are really the four things for an organization to look different.
- [11:58]
A
So when are companies, in creating this digital brain, I'm just curious, do they know, first of all, does it look like, do they have an idea and then create that brain? Just give us an idea of the journey. If I'm going to start a digital brain, where do I start and how do I get there?
- [12:13]
C
So a lot of companies have been taking steps forward in AI over the last few years. And in many ways, something which early movers did was they considered AI to be not that different from a digital transformation. And the recipe for digital transformation was to come up with use cases, proof of concepts, and then to learn from those, to deploy them, and to take advantage of it. AI is not that. AI is a general purpose technology. Once you have your data and your architecture to make decisions using your brain, from that data, you have infinite use cases. It's not just one single use case. And, and so those companies that are poised to most benefit are those that have recognized this isn't death by a hundred or Even I've seen 5,000 use cases, but it's reinventing their business, asking which processes change, how do I change the work, how do I get the right workforce to use across those? And how do I make the right decisions? All of that investment is complicated. It takes governance. But the leaders we're working with today and where I really see the exciting opportunities are really recognizing the importance of putting that digital architecture in place, not just for the use cases of yesterday, but for all the opportunities going forward.
- [13:32]
A
How about the workforce itself? What are the workforce needs and how do you make them AI ready with the necessary skills? And do you see some jobs remaining, some going away? What's the landscape for people?
- [13:46]
C
So from our research, 70% of executives cited that AI reskilling is a top priority for the next three years. Now this means that they really have to consider what are the skills of the future, how do I adapt my role and what do I take? I work with Accenture. And we've recognized for well over the last year to 18 months that our work is changing, the expectations of organizations that we partner with are changing, and we are building internally the mechanisms that to ensure we're bringing AI literate people with bespoke capabilities, not just in terms of using these tools, but in many cases building them and taking them further. I mentioned it earlier, the question that many people are going to have to grapple with long and hard is what does it take to be exceptional at working with and leading with AI? And I think what we're increasingly seeing is that judgment is more important now than ever. There was a really interesting bit of research that anthropic that makes Claude put out recently. And it definitely made me think, because what it showed was that if you give two workers access to the same AI tool, they do not get the same outcome, they do not get the same quality of outcome, and their incoming skill set knowledge, familiarity with the tool can actually lead to very different quality of insight that comes from that. In some cases, yes, AI will be a leveler. I think that goes without saying. But I think for those people, those workforces that lean in, there is a lot, there is a huge amount of differentiation that winners will come from.
- [15:23]
A
Last question for you. What's your advice to the company CEOs who are on the front line of on this AI journey? What should they be doing to get going and get ahead of the game,
- [15:33]
C
if you will, at all? You could have asked me this one first and would have probably taken most of the time that we had today just going through that. But let me build on a couple of things that we've discussed through this afternoon's discussion. The first was I mentioned that this is not just a technology question. If you reflect on everything we've talked around with how does IT change organizations, how does it change work, workforce decision making and structure, what you realize is that AI is really, it's an operating model bed. And first and foremost, those CEOs that see this as just the next few, few basis points or something to talk about in their investor communications perhaps are missing the real change which is happening, which is that AI is opening up an opportunity for a reinvention or a near total operating model Rethink the second pudding from our research, something that I do hope is thought about is that we have seen that those companies that have started on their AI journey or actually quite advanced in their AI journeys continue to overspend on technology relative to the total program. And on average, for every dollar that is invested in an AI program, $0.70 is spent on the technology itself, 30% is spent on changing the work, embedding it, changing the workforce, building those skills that we just talked about and making sure that change sticks. Those companies that have seen the best results flip that ratio. And so one third on technology, 2/3 on getting the results from it and changing the organization. And then the final one I would leave you with is that if you think about what I was talking about earlier around, a lot of AI is doing what we do today, but doing it for less inputs. We're again now seeing that change. And so outside of the energy industry, around 80% of CEOs that we surveyed saw AI primarily as being a driver of revenue and of growth.
- [17:26]
A
80%.
- [17:27]
C
80%.
- [17:28]
A
And in the energy industry, I don't
- [17:30]
C
know, we definitely should ask this group, but the majority haven't emphasized as much that AI is a source of new revenue, new opportunities. And throughout all my conversations at ciraweek, I'm still hearing questions around how can we get to product level carbon accounting? How can I bundle? How can I decouple the consumption and the cost of my goods? How could I monetize my data differently? How can I increase my recovery? All of these things open up new markets and new growth. And so part of what I'd like to leave this call with is to reflect that AI isn't just around efficiency, it's actually an upside. It's growth for us. And one of the key messages I think that I picked up today coming back to that energy is life and we need energy in abundance. And so if we can make AI part of solving for that, I think we're all doing the right thing.
- [18:22]
A
Yeah. First of all, David, thank you very much for this conversation. What I will reflect on is many companies are on the journey, but at different places in the journey and they all want to get there as soon as possible. But it's going to take some time depending on how committed. And as you said, it has to come from the top. And that's a very critical point for every company, every person who's listening to this podcast. Any final comments from you, Atul, thank
- [18:47]
C
you for spending the time this afternoon. This topic is one that's going to run and run. It's going to continue getting more and more exciting. And alongside of course, gas.
- [18:55]
A
Yes, we shouldn't forget gas.
- [18:56]
C
Cannot forget gas. AI is really shaping the narrative at ciraweek, has done for the last year or two and I think great to see that continuing.
- [19:03]
A
I would say that gas will be there for a while, but AI is going to be there for a much longer time. The gas crisis will hopefully go away soon. LNG and Natural Gas Crisis the David Rabley from Accenture, thank you very much for joining us for this conversation.
- [19:16]
C
Thank you, Atul.
- [19:18]
B
Thank you for joining us on the Sarah Week podcast to stay connected with the ideas driving change across energy and technology. Subscribe, share and rate this episode. It helps us get the word out. Let's continue having impactful conversations.
- [19:33]
A
I'm Atul Arya.
- [19:34]
B
Until the next time, the Sara Week
- [19:37]
D
Podcast with Atul Arya is brought to you by saraweek, the world's premier energy conference. Be part of the conversations moving the world of Energy Forward, March 8th through 12th, 2027 in Houston. Discover more@saraweek.com.