Breaking the Mold: How Unconventional Thinkers are Rewriting Energy’s Future

It’s often like Christmas at my job. Well, more specifically, AI Christmas for Energy. 

​As a Senior Product Manager at ThinkOnward, I help support and run data science challenges in the form of multi-month hackathons that draw talent from around the globe. It was a frosty winter morning when our team gathered to review one of the most exceptional AI solutions we’d ever seen for its use case—and this day promised to be even more exciting than usual. 

Our team receives novel AI models for complex energy problems from brilliant minds worldwide, and we have the privilege of exploring and evaluating them. In the past couple of years, we’ve evaluated hundreds of models applied towards energy disciplines like geophysics for oil and gas or carbon capture, geothermal energy, and electric power. 

These challenges often originate from clients whose domain experts have made significant progress but are seeking additional perspectives to break through remaining barriers. It’s a common predicament in today’s business world: the AI field is advancing at breakneck speed, leaving many companies struggling to harness its full potential due to limited resources and expertise. 

When a data science challenge on our platform closes, we get to open the “goodie bag” of models and try applying them to those complex energy problems. That day, as the team huddled around with steaming cups of coffee, I couldn’t help but feel a mix of excitement and curiosity. 

The challenge which just closed was particularly demanding—it required using a cutting-edge AI model (Meta’s Segment Anything Model) in an entirely new way: to automatically identify and map complex geological layers across an enormous dataset of over 9,000 3D seismic volumes. This was like asking an AI system designed for regular photos to interpret complex 3D X-rays of the Earth—crucial data that helps us find underground reservoirs, monitor carbon storage, and manage groundwater resources. What solutions would our global participants have cooked up this time? 

Our lead evaluator walked through the top models. While all performed well at the basic task—identifying underground structures from seismic data—one solution stood out dramatically. Not only did it use creative new techniques, but it could do something remarkable: it worked effectively across many different types of terrain and geological conditions. While most solutions excelled in specific scenarios, this one maintained its high performance regardless of the geological environment it encountered. 

It was clear the team weren’t using standard geoscience approaches, so we were keen to meet the minds behind this outstanding solution.  

A week later, we found ourselves on a video call with the team. It turned out that they were medical imaging professionals who spend their days analyzing MRIs and CT scans of patients, and who’d adapted their medical diagnostic techniques to our seismic data. These “outsiders” had seen connections that “insiders” may have missed, applying their cross-disciplinary knowledge in ways we hadn’t seen before. 

Their breakthrough illustrates a powerful lesson for any organization facing complex technical challenges: sometimes the most transformative solutions come not from industry experts, but from brilliant minds who can adapt knowledge from entirely different fields.

In our experience at ThinkOnward, we’ve observed a phenomenon: the most innovative solutions to complex energy problems often come from unexpected sources. These breakthrough ideas aren’t always from industry insiders or domain experts, but from individuals we might call “outsider polymaths”. 

This is true even for highly specialized problems like geoscience, which might seem out of reach for non-domain experts. These outsider polymaths bring a unique blend of cross-disciplinary knowledge and fresh perspectives to the table. They’re able to see connections and possibilities that those deeply entrenched in the industry might overlook. By applying insights from seemingly unrelated fields, they can sometimes crack the code on long-standing problems that have stumped traditional experts. This is a significant finding because businesses conventionally look to domain experts to advance a field. By doing so, they may be leaving improvements of orders of magnitude, and more effective or game-changing solutions to their problems, on the table. Financially, (in our experience comparing the two) this difference sometimes represents millions of dollars. The impact of outsider polymaths extends far beyond financial benefits. Their unique approach to problem-solving offers a fresh perspective that can revolutionize entire industries. By drawing from diverse fields of knowledge, these innovative thinkers bring a multifaceted approach to complex challenges. ​​​In the following sections, we’ll explore the key characteristics that make outsider polymaths so effective, and examine real-world examples of how their cross-disciplinary expertise leads to groundbreaking solutions in the energy sector. From integrating seemingly unrelated fields to asking unconventional questions, these polymaths demonstrate the power of thinking beyond traditional boundaries. 

The Essential Foundation: Why Domain Expertise Remains Critical 

Before we discuss examples of cross-disciplinary breakthroughs, it's crucial to establish an important truth: outsider polymaths succeed precisely because domain experts have built the foundation they're standing on.    The medical imaging team couldn't have solved our seismic challenge without: 

• Decades of geoscience research that identified what features matter in seismic data 

• Domain experts who framed the problem clearly and provided labeled training data 

• Industry knowledge that defined success criteria and validated the results   

In our experience at ThinkOnward, the most successful outcomes occur when outsider polymaths can stand on the shoulders of domain experts. We provide labeled datasets, comprehensive reference materials, and problem frameworks. The outsider brings fresh techniques and novel approaches; the domain expert's knowledge provides context, validates assumptions, and ensures the solution addresses the real-world problem.   The goal isn't to replace domain experts—rather it's to bring in fresh perspectives to build upon the foundation of knowledge that domain experts have established. When we provide outsider polymaths with high-quality data and domain knowledge, we create the conditions for breakthrough innovations that leverage the best of both worlds. 

Outsider Polymaths Integrate Expertise Across Domains 

While deep domain expertise is crucial for solving specific technical problems, many of the most complex challenges require input from multiple fields. The most successful outsider polymaths aren’t just people from outside the industry—they’re experts who have mastered complex problem-solving in another field and can creatively apply that knowledge to energy challenges. They may also excel at bringing together specialists to tackle multifaceted issues. 

As we saw with the medical imaging experts who cracked our seismic problem, these polymaths excel at recognizing how techniques from their own domain can unlock solutions in seemingly unrelated fields. 

From MRIs to Earth’s Layers: How Medical Minds Cracked the Seismic Code 

Back to the ThinkOnward geoscience challenge in the introduction—it involved 3D seismic data, which has numerous uses such as reservoir identification and monitoring for CO2 sequestration, oil and gas exploration, and groundwater identification and management. 

The challenge required participants to use the seismic data to identify and map layers of the Earth, in order to interpret geologic features reasonably. 

​​​The standout solution was from a two-person team—a medical student and machine learning engineer—both working at a hospital. In their day jobs, they worked on segmentation for huge datasets of 3D medical images to improve diagnostic accuracy and enable more precise treatment planning. 

Their computer vision skills in managing 3D and multimodal medical data, such as ensuring spatial context across multiple slices of data, were highly transferrable to 3D seismic volumes. 

Medical imaging often involves noisy, low-contrast data (e.g., MRI, CT scans) that require robust preprocessing techniques such as denoising, normalization, and contrast enhancement. Seismic imaging also suffers from noise and low signal-to-noise ratios, so the team’s skills in handling these preprocessing tasks for the medical industry were transferrable to a complex geoscientific problem. 

Despite having zero experience in geoscience, the team developed an impressive model for seismic interpretation with their skills from medical diagnostics. 

Outsider Polymaths Catalyze Innovation by Connecting Dots Others Might Miss 

Domain experts sometimes operate within the confines of their field's established knowledge. When working alongside these experts, outsiders can contribute complementary perspectives—while the expert ensures technical validity and practical applicability, the outsider may spot connections to techniques from other fields that haven't yet been tried. 

Streaming Success: How Video Buffering Solved Oil Well Mysteries 

​​​Imagine an oil well as a giant, underground YouTube video. Sounds crazy, right? But that’s exactly the kind of out-of-the-box thinking that led to a breakthrough in another ThinkOnward challenge. 

We asked participants to build a model predicting whether an oil and gas well was flowing or shut in—a crucial task for optimizing operations, improving safety, and boosting efficiency. They were provided with bottom hole pressure data which hold information about patterns, anomalies, and potential opportunities. 

Enter our unexpected hero: a machine learning engineer who’d spent years battling the dreaded “buffering” icon for a major video sharing platform. Our challenge winner saw a striking similarity between managing video streaming congestion and monitoring oil flow. 

In the world of streaming, engineers constantly analyze packet loss, buffer levels, and bandwidth fluctuations to predict and prevent congestion. Our video virtuoso realized that predicting well flow wasn’t so different. Instead of packets and buffers, they were dealing with pressure and flow rates. But the underlying principle was the same: detecting and managing “congestion” in the system. 

By adapting algorithms designed to keep your favorite shows streaming smoothly, our winner created a model that could predict well behavior with remarkable accuracy. 

Outsider Polymaths Ask Different Questions and Redefine Problems 

Domain experts are trained to ask certain types of questions within their field, but polymaths can approach problems from entirely different angles. This can lead to redefining the problem in ways that make it easier to solve or open up entirely new avenues of exploration. 

Cosmic Debris Meets Cell Chemistry: Space Expert Energizes Battery ​​Tech 

The ThinkOnward challenge Future in Charges focused on improving the efficiency of battery technology by modeling electrochemical processes, specifically through the understanding and solving of the Poisson-Boltzmann Equation (PBE). The competition was broken into three sequential challenges, each increasing in complexity, to develop computationally efficient solutions for modeling electric double layers—key structures in electrochemical systems like batteries. The winning solution came from a researcher with vast experience in the space industry, particularly in modeling space junk and debris. This unexpected crossover between celestial mechanics and electrochemistry exemplifies the power of outsider polymathic thinking. While we don’t know exactly what questions this space expert posed during the challenge, we can imagine how their unique perspective might have led them to ask very different questions to those of a battery expert. A space junk expert is likely to explore spatial distribution and flow differently to the battery expert. Space junk experts are used to tracking the spatial distribution of objects and how forces (e.g. gravity) affect their trajectories. In this competition, they might have asked, “How does the charge distribution propagate spatially within and across surfaces over time?” or “What are the spatial patterns of charge accumulation in complex geometries, and how do they interact with the boundaries?” This could lead to redefining the problem from a static 3D model into one that explores charge flow as an orbital-like phenomenon, where field lines resemble paths akin to space trajectories. In comparison, a battery expert would likely focus more on how charges are stored and transported at a molecular level, considering the impact of specific materials and chemical reactions within the battery. Their questions might be more material-centric, like “How do different battery chemistries affect the electric double layer?” Generally, a space junk expert might ask broader, more system-level and dynamic questions, focusing on computational efficiency, uncertainty management, and large-scale interactions. They may redefine the challenge by applying concepts from orbital mechanics, fluid dynamics, and surrogate modeling to simplify and scale up solutions. In contrast, a battery expert is more likely to focus on chemical-specific optimizations, material interactions, and detailed electrochemical phenomena, working within well-understood battery frameworks. Through asking different questions, the space junk expert’s unconventional perspective ultimately led to creative, cross-disciplinary approaches that offered new insights into battery modeling. 

Putting Polymathic Thinking into Practice: A Business Guide 

The examples above demonstrate the power of cross-disciplinary thinking. But how can businesses harness this potential? Outsider polymathic thinking enables leaders to draw on knowledge from multiple domains, allowing for more creative and innovative solutions to complex business problems. By integrating diverse perspectives, a company can quickly adapt to changing market conditions, anticipate industry shifts, and seize opportunities that competitors might miss. How can businesses seek out and nurture outsider polymath talent for their needs, or even to develop a competitive advantage? Here are some ways I’ve seen work well for problems requiring innovative solutions. 

1. Ideate with people from outside the industry  In ideation sessions for tackling such problems, bring in people from outside of your industry. If they need some foundational understanding of the problem, send them pre-reads or provide a 101 crash course to the topic. 

2. Nurture polymathic creativity within the company  We regularly encourage team members to share an observation or learning that sparked ideas, with the caveat that it must be from outside of the energy industry (our industry). 

3. Partner with an innovation platform  Platforms like ThinkOnward can help you frame your problem into a format which optimizes for attracting divergent approaches from a global talent community, while ensuring that solutions meet your goals. ThinkOnward also sources, vets and curates talent so that they are available “on demand”. 

Like doing anything new, trying these may take some trial and error. For example, it may take a few tries to understand what alternative industries you may be looking to get new ideas from (such as medical imaging concepts being applied to 3D seismic imaging problems). 

Modern Challenges Demand Diverse Perspectives 

Many of today’s most pressing problems—like climate change, pandemics, and economic inequality—are complex and interconnected. These problems cannot be solved by a single domain expert. Instead, outsider polymathic thinkers who understand the intersection of various fields (science, policy, economics) are better equipped to approach these global issues holistically. While polymathic thinking offers significant advantages for complex, interdisciplinary problems, it’s important to note that domain expertise remains crucial for certain tasks. For instance, data labeling for model training often requires in-domain experts who can accurately identify specific elements, such as distinguishing a tumor in a medical image or a seismic fault line in geological data. The ideal approach often involves combining outsider polymathic perspectives with domain-specific knowledge. For any ThinkOnward challenge, we usually provide some foundational understanding to start the participants off. In addition, we provide office hours and a forum where participants can pose questions to domain experts. Facilitating such collaboration has helped outsiders to quickly grasp an unfamiliar domain while being unencumbered by legacy thinking within the domain.  

The Future of Innovation: Embracing the Polymath Advantage 

Back at the job, AI Christmas for Energy continues. A client of ours executed strategies 1 and 3 above—they contracted two ThinkOnward geoscience challenge winners with no geoscience background, to ideate and then work on a complex tomography inversion problem. They created significant new efficiencies which in-domain experts did not consider for years. 

The energy industry has no shortage of complex problems waiting to be solved differently and better. By creating partnerships between domain experts and cross-disciplinary thinkers, we've tapped into new ways of thinking to uncover unexpected new solutions that neither group could have achieved alone. Some of these have been valued by clients at millions of dollars in ROI. 

In sharing our experience, I hope this will encourage you to consider how you might harness outsider polymathic thinking for your organization. By implementing these strategies, you might just nurture the next Da Vinci within your own organization, advancing new frontiers for humanity.