In the dynamic world of oil and gas, innovation has always been the bedrock of competitive advantage. From the advent of hydraulic fracturing to advanced seismic imaging, our industry thrives on technological breakthroughs that redefine efficiency and profitability. Now, a seismic shift emanating from the software sector offers a prescient glimpse into the profound, transformative power of artificial intelligence, a disruption poised to reshape every facet of the global energy landscape, including upstream, midstream, and downstream operations.
For decades, software engineers occupied a privileged position, their expertise driving the digital revolution. Today, these highly skilled professionals are witnessing an unprecedented redefinition of their roles, propelled by a rapid evolution in AI capabilities. Late last year, major players like OpenAI, Anthropic, and Google unleashed new AI models that dramatically enhanced coding tools. Almost overnight, AI agents gained the ability to tackle complex programming tasks, functions that previously demanded years of human mastery. This swift technological advancement serves as a powerful harbinger for the oil and gas sector, illustrating how quickly AI can move from conceptual promise to tangible, operational impact.
Consider the recent experience of Amy Surrett, an engineer in Greenville, South Carolina. In January, she leveraged Anthropic’s Claude Code to develop a payment feature for a client. A project that would typically consume two to three days of manual effort was completed in just over an hour. “It felt like the point of no return,” she observed. “This industry is not going to be the same. My job is not going to be the same.” This sentiment resonates deeply with the pace of change we anticipate in our own industry, where AI-driven optimization could slash project timelines and capital expenditure in exploration and production.
Andrej Karpathy, a prominent research scientist now with Anthropic and formerly with OpenAI, highlighted this sudden acceleration. In a February post, he described the radical transformation, noting that prior to December, coding agents were largely ineffective. Then, almost instantaneously, a “takeoff” occurred. Fast forward to June, and the software engineering profession, which employs tens of millions globally, finds itself in a full-blown reckoning. AI fuels fears of job displacement, introduces new terminology like “tokenmaxxing” (optimizing AI agent interactions), and attracts hundreds of billions in fresh investment. At tech giants like Google, AI is now responsible for generating as much as 75% of the company’s code, a statistic that should command the attention of every O&G investor tracking operational efficiency gains.
Software developers, perhaps more than any other white-collar profession, represent the vanguard of AI integration into the workforce. Their experience provides invaluable lessons as this technology prepares to disrupt other sectors. Given the well-defined, logic-driven nature of coding, it has proven particularly susceptible to AI augmentation. The insights gleaned from this “Great Coding Reset” hold broad relevance, signaling how AI could fundamentally reshape everything from reservoir modeling and drilling automation to supply chain logistics and predictive maintenance in the oil and gas value chain.
The Convergence of Innovation and Disruption
The software industry has always been characterized by cycles of reinvention. From the birth of personal computing in the 1970s, demanding talent for operating systems and programming languages, to the internet boom, then the mobile revolution, each era reshaped the landscape. Today, the emergence of “no-code” platforms like Lovable and Base44 allows app development without traditional coding, further compressing the timeline for innovation. AI now accelerates this pace of change to an unprecedented degree, mirroring the intense periods of technological disruption the oil and gas industry has navigated with the rise of unconventional resources or deepwater exploration.
For many professionals, this velocity of change is both exhilarating and unsettling. Kent Dodds, who left PayPal in 2019 to educate software engineers, experienced this first-hand. He sought to create a tool for his students to securely download videos offline. In January, an AI-assisted coding tool, Cursor, “nailed it on first try,” he recounted, effectively eliminating weeks of development work in a single morning. “That was my first existential crisis,” Dodds admitted. His experience underscores how quickly traditional workflow can be subsumed by AI, challenging long-held assumptions about value creation and human contribution. As these models continue their relentless improvement, Dodds believes they are “behaving a lot more like a regular software developer,” with no discernible ceiling in sight for their capabilities.
AI as the New Operational Director
At a recent AI Engineer Europe conference in London, the prevailing sentiment was one of electrifying transformation. Ryan Lopopolo, a member of OpenAI’s technical staff, declared from the stage, “In the last six months, we have seen coding agents take over the world.” He posited that the software engineer’s role is rapidly evolving into that of an “agent supervisor.” Alex Ponomarev, founder of Volt, a software development agency, described the shift from “synchronous” AI interaction, where humans constantly guide the AI, to a more autonomous model. In the oil and gas context, this translates to AI-powered drilling rigs, autonomous subsea inspection systems, or smart fields requiring human oversight and strategic direction rather than minute-by-minute manual intervention.
This does not, however, imply a uniform reduction in workload. Some engineers voice frustration with the necessity of refining AI-generated code or rectifying suboptimal applications produced by less tech-savvy colleagues. The daily routine for developers is changing, with more time allocated to crafting detailed specifications for AI agents and managing “token limits” – essentially, the AI’s processing capacity. As Danial Qureshi, a Toronto-based software developer, previously noted, “It’s basically not even worth my time to be manually writing code when I can have something like Claude doing it for me.” This operational shift, focusing on higher-level direction and quality control, presents both efficiency gains and new challenges for workforce integration.
Strategic Vision Over Manual Execution
The profound capability of AI raises a crucial question for investors: if everyone has access to this extraordinary power, where does competitive advantage lie? The answer, increasingly, points towards distinctly human attributes: strategic vision, critical judgment, and innovative problem-solving. Kent Dodds, observing a “drastic decline” in student queries thanks to AI’s ability to provide rapid answers, has pivoted his curriculum. His new focus is “product engineering” – emphasizing what to build, not how to build it. The true value of human expertise, he argues, is in identifying the right problems to solve, assessing benefits and drawbacks, and understanding what genuinely serves the user or, in our context, the operational imperative.
“I’m teaching the last skill that the last software engineer needs to have,” Dodds asserts, highlighting the enduring importance of strategic acumen. Amy Surrett, who graduated with a Bachelor’s in software development just months before ChatGPT’s debut, now estimates AI tools generate 80-90% of her code, up from 5-10% a year ago. She acknowledges the “double-edged sword” – increased output with less direct coding – but remains confident in her human edge. “I know patterns of writing good software that someone who downloaded Claude Code last week won’t know.” Moreover, she hones “soft skills AI can’t really replace,” such as client communication, creative problem-solving, and discerning client needs – all critical for successful project deployment in any industry.
While AI is frequently cited in corporate layoffs, a nuanced view of the labor market reveals a modest uptick in software engineering job postings. This suggests that while certain roles may diminish, the demand for human professionals capable of overseeing, guiding, and refining AI outputs is growing. Jason Young, a 30-year engineering veteran, initially felt “very threatened” by these autonomous agents. However, extensive interaction with AI has solidified his conviction that human judgment remains indispensable. For Young, the core of engineering lies in comprehending the problem, not merely generating lines of code. “The writing of text — that isn’t what being a software engineer is,” Young states emphatically. “Anyone who thinks otherwise has a wild misunderstanding of software engineering.” This perspective offers crucial guidance for oil and gas executives and investors: AI is a powerful tool for execution, but human intellect will continue to drive strategic direction, innovation, and ultimately, value creation in the energy transition.