AI & Digital Twins: Fear and a Bold Vision

Four Terrifying Reasons AI and Digital Twins Are ​Tied at the Hip

The Design Futures Council conference in La Jolla last week wasn’t just a meeting of great minds—it was a wake-up call. It compelled us to face some of the most pressing consequential questions of our time, especially the relationship between AI, energy, and the built environment.

Although I already knew that AI was a significant issue in terms of energy use, Caroline Golin, Google's Global Head of Energy Market Development and Innovation, highlighted the staggering energy demands of AI systems and vividly painted a picture of the interconnected crises of politics, resources, and crumbling infrastructure. Her message was clear: AI is rewriting the energy landscape, and the way we currently operate is unsustainable. 
​

This isn’t just a feel-good tree-hugger challenge; it’s the future of everything—our technology, energy, and built environment depend on addressing these issues now. Her narrative solidified the 'why'—why we, as an industry, must change now. Her presentation outlined the urgent challenges ahead and set the stage for rethinking how the built and natural environment, technology, and energy systems need to work together.

Source: Masanet et al. (2020), Cisco, IEA, Goldman Sachs Research The data center power demand for 2023 is an estimate.

Listening to Caroline reinforced my long-standing belief that Digital Twins are key to solving these challenges. While she masterfully outlined the problem, I saw how Digital Twins could fundamentally transform how we manage energy and infrastructure. After Caroline’s session, I presented with Paul Doherty on “Making Design Ongoing and Defensible - What’s at Stake?” and now, in this paper, we want to go even further.
​
Her narrative solidified the 'why'—why we, as an industry, must change now. Here, we outline a bold proposal to tackle these challenges head-on: transforming buildings into Virtual Power Plants, powered by Digital Twins, to drive energy resilience.

The Modern Silk Road—the worldwide network of interconnected systems and data—is at the heart of today’s geopolitical and economic landscape. Caroline’s presentation painted a picture of the immense energy challenges tied to AI, breaking them down into four distinct issues:

Reason 1: Unprecedented Energy Spikes 
AI is valuable but comes with a cost; for example, Google’s AI systems now consume energy equivalent to the entire state of Idaho, with projections suggesting this will soon rival the energy consumption of Washington state, and it will keep climbing.

As AI systems like those from Google and OpenAI continue to consume vast amounts of energy, the strain on our power grids intensifies. This surge in demand isn't just a corporate concern—it has tangible repercussions for everyday consumers. A report by the Jack Kemp Foundation projects that by 2029, electricity bills for consumers and small businesses could rise by up to 70% due to the energy consumption of AI data centers. While tech giants may absorb these costs, the average household and local businesses will likely bear the financial burden, leading to significant increases in monthly utility bills.

Reason 2: Systemic Infrastructure Limitations 
As Caroline put it, Our utility grid was designed for a different era—fossil fuels, centralized power plants, and predictable demand.  We simply are not ready politically, socially, and operationally for the explosive growth of AI-driven consumption. 

Reason 3: Disrupted Sustainability Goals 
As Caroline said, no amount of sustainability solutions built on the old way will solve the challenge—it’s simply not achievable. 

Reason 4: The Geopolitical Arms Race 
China trails behind the United States in AI advancements, but this is not solely an AI race—it’s a broader competition with profound implications for economic, technological, and energy security. Caroline underscored that if U.S. companies and policymakers fail to lead, not only in AI development but also in the construction of critical infrastructure like data centers, these capabilities will emerge elsewhere, most likely in China. This scenario would not only intensify global energy challenges but also shift significant geopolitical power.

While Caroline’s presentation laid bare the challenges, it also highlighted a path forward: Virtual Power Plants (VPPs). In her opinion, VPPs are not just an option—they are essential. This approach integrates distributed energy resources like solar panels, smart batteries, and renewable energy systems, coordinated through advanced technology, to meet the demands of a rapidly evolving energy landscape.

---

A Bold Proposal: Transforming Buildings into Virtual Power Plants

Our proposal is to connect VPP systems via digital twins to ensure all buildings are not only energy efficient but active contributors to grid stability. Digital Twins enable real-time optimization, helping VPPs balance energy production and demand while reducing reliance on fossil fuels. We simply can not continue to consume energy the way we have been without implementing measures that ensure we are optimizing every bit of its use.

While Caroline’s presentation laid bare the challenges, it also sparked ideas about solutions. Digital Twins—virtual replicas of physical assets—represent a critical opportunity to address the energy challenges faced by all types of facilities, not just those tied to AI. Whether it’s an office building, hospital, school, or AI-driven data center, Digital Twins are key to transforming energy consumption into optimized, efficient processes.

Digital Twins enable real-time energy optimization, tracking energy use across systems and adjusting to improve efficiency. For example, during periods of high demand, information from Digital Twins can help synchronize renewable energy sources like solar and wind with grid needs, ensuring a balanced energy flow. Additionally, they facilitate system integration, breaking down silos by connecting power grids, AI-driven data centers, and other networks, creating a seamless flow of information and control.
​
What excites us most about Digital Twins is their scalability and flexibility. They evolve with technological advancements, learning and adapting as demands change. This ability to respond dynamically makes them indispensable for tackling the challenges Caroline described, particularly at the scale required by AI’s energy trajectory.

A BIMStorm brings together an international team of owners, architects, engineers, and facility managers to tackle specific design problems using BIM and Digital Twins. This ongoing initiative, started in 2008, enables teams to collaborate online, eliminating the need for travel or large meeting spaces, and disband as quickly as they form.

​BIMStorm: Building Smarter, Powering ResilienceBuilding on these ideas, we propose we shift BIMStorm from a conceptual exercise to an actionable initiative. 
BIMStorm has tackled  Microgrids in 2022 and AI-driven Facilities and Assets in 2023-2024, demonstrating how collaborative innovation can solve complex problems. This year, we take the next logical step: integrating VPPs to address energy resilience at scale.
​
Imagine buildings generating energy on-site with solar panels or geothermal systems and storing excess energy in smart batteries. These systems, integrated with the grid via Digital Twins, could stabilize energy flows while reducing reliance on fossil fuels. By leveraging AI, BIMStorm will demonstrate how the built environment can evolve into an active energy solution rather than a passive consumer.

As we face the aftermath of recent devastating fires in Los Angeles, this challenge provides an opportunity to rebuild smarter rather than repeating the same mistakes. Instead of simply replacing damaged infrastructure, we can implement smarter systems and technologies that prioritize resilience and sustainability.

As architects, engineers, builders, and owners, we often celebrate LEED plaques on our buildings. While commendable, they fall short unless tied to the interconnected networks and Digital Twins that Caroline described. This is the moment to move from aspiration to action. The technology exists; we simply need the collective will to implement it.
​
We invite all forward-thinking leaders—owners, designers, and technologists—to join us. Together, we can transform how infrastructure is designed, built, and operated to power a more resilient future.
​
Conclusion:
Caroline’s presentation left us with a stark reality: there’s no turning back from the path AI has set us on. But as terrifying as these challenges are, they also present an incredible opportunity. By tying AI and Digital Twins together, we can turn these energy crises into a catalyst for systemic change. Let’s ensure the questions we’re asking today lead us to the answers that will shape a more sustainable tomorrow.

The Irony is Undeniable
​AI, while driving a staggering increase in energy demand, also presents the very tools we need to address the crisis it has intensified. AI forces us to rethink energy management at a pace far beyond human capability. Digital Twins, in tandem with AI, offer an actionable path forward. They empower us to achieve the level of real-time optimization, coordination, and resilience required to meet these demands sustainably.

While the energy crisis feels daunting, it also serves as a powerful motivator. With AI shaping an inevitable trajectory, the choice is no longer about whether to adapt but how quickly we can mobilize the tools at our disposal to move faster than ever. By harnessing AI and Digital Twins, we can not only mitigate the challenges of today but turn this crisis into an opportunity to build a more sustainable, resilient future.

We work with the digital transformation of many amazing owners, with many amazing LEED Certified buildings. But there is something missing not only with these owners, but with the vast majority of LEED buildings.

Real Owners, Real Achievements: Bridging the Gap
The building industry has made remarkable strides in sustainability, driven by the collective efforts of architects, engineers, facility managers, and other stakeholders. Across the globe, advancements like LEED certifications have set benchmarks for energy efficiency and environmental responsibility.

Today, there are:

• Over 195,000 LEED projects
• Spanning 186 countries
• Encompassing 29 billion square feet of certified space.

These projects are designed to save more than 120 million metric tons of CO2 emissions, highlighting the industry's commitment to combating climate change. But that is not enough.

Despite these successes, the industry still faces significant challenges. Buildings' current deliverables and data often act as liabilities rather than assets, contributing to inefficiencies and wasted energy. Caroline’s presentation underscored this reality, emphasizing that the problems cannot be solved with incremental improvements—they require a complete digital transformation.
​
By embracing Digital Twin technology, the industry can turn buildings into valuable, high-performing assets that drive performance, resilience, and sustainability. Digital Twins enable impressive designs to evolve into systems that actively contribute to energy efficiency and environmental stewardship, bridging the gap between aspiration and operational performance. The path forward is clear: the industry must prioritize digital transformation to unlock the full potential of the built environment.

1st in a series of blog posts:
1. Four Terrifying Reasons AI and Digital Twins Are ​Tied at the Hip
2. Yelling Fire in a Crowded City
3. Why Are Smart Buildings and Cities…Dumb?

---

Kimon Onuma, FAIA
A recognized leader in BIM and Digital Twins, Kimon has over three decades of experience Building Informed Environments. His work focuses on open standards, interoperability, and innovative lifecycle management solutions, including the ONUMA System, BIMgenie, and BIMStorm. Kimon also leads the development of industry standards, contributing to the National Institute of Building Sciences' Digital Twin Integration Subcommittee and advancing AI integration in building processes.