Edgeworks Revolution: Bringing AI Compute to Where Power Lives

Energy News Beat

Following our article yesterday, here is a discussion on “Data Center Solutions Done Right,” as we say in Texas.

In this episode of the Energy News Beat Podcast, host Stu Turley sits down with Jon Brewton, CEO and Founder of Data2, and Kyle Koss, President of ATCO Ventures, to explore a critical challenge facing the AI revolution: infrastructure. As artificial intelligence demand skyrockets—with projections suggesting a 5,000% increase in compute requirements over the next five years—traditional mega data centers are hitting a wall.

The solution? ATCO Edgeworks, a groundbreaking modular edge computing platform that brings AI infrastructure closer to power sources rather than centralizing it in massive facilities. This conversation delves into how distributed, modular data centers can democratize AI access for mid-sized companies, solve environmental and community concerns, and create a more resilient computing landscape.

With insights into explainable AI, military applications, and real-world use cases from the energy sector, this episode reveals why the future of AI infrastructure isn’t about building bigger—it’s about building smarter and closer to home.

In my article yesterday, “The backlash is real, bipartisan, and already killing projects,” the tagline was “Data Centers Done Right are a good thing – Done Wrong, and we get problems. Well, today’s podcast really emphasizes how to do Data Centers right. Build them closer to the power sources, and don’t leave consumers stuck with billions in unnecessary transmission lines.

Connect with Kyle on his LinkedIn: https://www.linkedin.com/in/kyle-koss-3bbb1927/

Check out ATCO here: https://www.atco.com/en-ca.html

Connect with Jon on his LinkedIn: https://www.linkedin.com/in/jon-brewton-datasquared/

Data 2 website: https://www.data2.ai/

Watch this video on YouTube.

1. Distributed Edge Computing Infrastructure

The core focus is on ATCO Edgeworks, a modular data center solution designed to bring compute closer to power sources rather than centralizing it. This addresses the massive infrastructure bottleneck created by AI demand, which requires enormous capital investments and long lead times (3-5 years) for traditional mega data centers.

2. AI Infrastructure Bottleneck & Power Constraints

The guests discuss how AI demand is exploding exponentially, with Goldman Sachs projecting a 5,000% increase in compute demand over the next half decade. The challenge isn’t just computing power—it’s the massive power requirements, capital commitments, and infrastructure constraints that only a handful of mega-companies can afford.

3. Democratizing AI Access

A key theme is making AI infrastructure accessible to mid-tier companies (the “mom and pop” oil and gas operators) who can’t afford billion-dollar data center commitments. Edgeworks enables smaller organizations to deploy AI capabilities without massive upfront capital or 15-20 year power contracts.

4. Environmental & Community Impact (NIMBY Solutions)

The discussion addresses concerns about data centers’ environmental footprint—water usage, cooling requirements, power grid strain, and community opposition. Modular edge computing reduces these impacts by distributing load across smaller, localized units rather than massive centralized facilities.

5. Data2’s Explainable AI & Trust Layer

Jon Brewton emphasizes the importance of AI validation and explainability. His company Data2 has patented technology that adds a “trust layer” to AI systems, making decisions auditable and trustworthy. They cite a real example where their platform discovered $98 million in hidden fraud risk.

6. Military & Defense Applications

The guests discuss edge compute’s critical role in defense—portable, distributed systems that can be deployed in the field, reducing vulnerability from centralized targets and enabling real-time AI decision-making for warfighters.

7. Flexible, Modular Design Benefits

The Edgeworks solution is built entirely on modular architecture (hardware and software), allowing customers to:

Scale up or down based on demand
Use existing power sources (natural gas, solar, flared gas)
Avoid single points of failure
Reduce unnecessary data movement

8. Real-World Use Cases

Examples discussed include:

Oil and gas companies monetizing flared natural gas to power compute
Bitcoin mining operations pivoting to AI workloads
Industrial sites leveraging existing infrastructure for edge compute
Military deployment scenarios

9. Market Timing & Competitive Advantage

The guests argue this is the right moment for distributed edge computing because: AI demand is rising exponentially, power availability is constrained, and capital requirements for centralized infrastructure are prohibitive. This creates market demand for alternative architectures.

10. Data Sovereignty & Security

Partnerships with encryption providers (like XQ) ensure quantum-encrypted, zero-trust security—critical for government and military applications, and for protecting sensitive data in distributed environments.

The overarching narrative: Traditional mega data centers are hitting infrastructure limits. Edgeworks offers a paradigm shift by bringing compute to where power exists, enabling broader participation in AI, reducing environmental impact, and creating a more resilient, distributed compute market.

It was fun visiting with Kyle, and as a Canadian Company with manufacturing in the States, it makes it easier for them to keep rolling through the trade imbalances.

Thanks to Jon and Kyle for a fun discussion, and let us know if you have any data center concerns or need help with AI and implementation at scale.