As we set sail into 2026, I find myself reflecting on how much has changed over the past three years. The steady drumbeat of “peak oil” and “peak demand” continues to echo across the industry, yet global demand still rises in step with production. U.S. production was 13.78 million barrels of oil per day at the end of the year! The long-anticipated “oil glut” – those millions of barrels floating toward market – has yet to materialize onshore. Many experts believe this will occur; the question is when. In a similar manner, attention is turning toward an expected LNG surplus projected for 2027–28. While new projects continue to advance, associated gas production from the Lower 48 could decline significantly if oil prices remain low. This could potentially impact future gas supply. Similarly, large-scale LNG projects are also prone to delays and increased costs which could greatly affect the timing of future supply.
The Hype and Promise of AI
Electric vehicles were once an anticipated growth sector in the auto industry. Yet we’re now witnessing a precipitous decline in EV sales. The solar panel power sector is predicted to continue its meteoric rise. But the reality of base-load power demand and the loss of subsidies will certainly impact on the growth of this industry. And then there’s artificial intelligence – the technology many claim will change everything. I’m not sure it will change absolutely everything, but I do see remarkable potential in its applications.
In January, I had the privilege of attending the International Petroleum Technology Conference in Dubai. It was a short, two-day meeting focused on a massive topic: how AI is being applied within petroleum technology. I’ll admit, I wasn’t certain how much I would enjoy the event, but the program was excellent and I left with valuable insights into how companies are adopting AI, how geoscientists are integrating it into their workflows and how much I need to learn.
The discussions were broad-ranging, and it was clear that AI applications in geoscience are still evolving. On a personal level, I realized just how far down the learning curve I am, and the importance of investing time to understand how AI can enhance my daily work, improve efficiency, and shorten cycle times. Like any new technology, there will be frustrations and challenges, and I expect to rely on both my colleagues and my kids for a bit of support along the way.
One key takeaway from the conference was what I did not hear: no one suggested that AI will replace people. That, perhaps, was the most important lesson. Industry leaders emphasized that AI is a tool – a powerful one – but still dependent on human guidance, teaching, and evaluation.
I also learned a new term:
“hallucination.” A young professional from Microsoft, who happened to sit next to me on the airplane, explained that AI sometimes “hallucinates” when it encounters gaps in data or logic, producing results that appear confident but are simply inaccurate. It’s a fascinating concept and, at its core, just another form of bad data interpretation. Maybe Timothy Leary isn’t dead after all.
Personally, I have experienced what I call “AI bias.” I have asked for revisions and edits on some of my writing and have discovered the entire context was changed, making the opposite point. It might be an indication of the programming and as the Great and Powerful Oz said, “Pay no attention to that man behind the curtain.”
The AI race seems to have taken on a life of its own, with countless voices predicting how it will transform our world. There’s truth in that – AI will reshape many aspects of our work; but it’s important to remember that we, not technology, control the pace and direction of its evolution in its application to geoscience and the subsurface.
What Does AI Mean for the Geoscientist?
As geoscientists, we stand at an exciting frontier. AI gives us the ability to streamline workflows, evaluate multiple scenarios simultaneously, and improve both decision quality and interpretation accuracy. Over the past week, I’ve realized that nearly everyone in our industry is grappling with how to best apply AI and large language models to their work. The driving force behind it all is computational power – the “energy” that enables us to perform analyses thousands, even millions, of times faster. This will underpin the next wave of progress in reservoir modeling, production optimization, drilling prediction, and operational efficiency – all of which can potentially help reduce cost and uncertainty.
As I mentioned last month, AI also has tremendous potential as a teaching bridge for the next generation of geoscientists. It can rapidly generate summaries of technical topics and link directly to research and publications relevant to the questions being asked. Used wisely, it can enhance how we learn, share, and innovate as an industry.