From Cloud to Device Edge, AI Deployment Is Now a Sustainability Decision for Enterprises - Episode 126

From Cloud to Device Edge, AI Deployment Is Now a Sustainability Decision for Enterprises - Episode 126

Let us know how we're doing - text us feedback or thoughts on episode content [https://www.buzzsprout.com/2064554/fan_mail/new] Most companies think of AI's carbon footprint as a data center problem. But the infrastructure powering your AI queries is rapidly diversifying — and each deployment model carries a very different sustainability profile. In this episode, Paul breaks down five emerging AI infrastructure options and what they mean for corporate emissions accounting. Paul makes the case that as AI deployment shifts away from traditional cloud infrastructure, the CSO has to be at the table with the CIO and CFO. The deployment decision is one of the most consequential choices a company can make for its GHG profile — and right now, most organizations are making it blind to the emissions implications. Follow Paul on LinkedIn [https://www.linkedin.com/in/schusterpaul/].

Why Data Centers Are Raising Your Electricity Bill — And "Bring Your Own Power" Isn't Fixing It - Episode 125

Let us know how we're doing - text us feedback or thoughts on episode content [https://www.buzzsprout.com/2064554/fan_mail/new] Your electricity bill is going up — and data centers are a big reason why. In this episode, Paul cuts through the PR and misinformation to explain exactly how the AI data center boom is driving residential utility rate increases, even when hyperscalers claim to "bring their own power." Paul breaks down four ways data centers are straining the grid and socializing costs onto ratepayers: backup power redundancy requirements, exploding capacity market prices, transmission infrastructure costs, and wholesale price contagion that can send electricity costs soaring for communities near new data center sites. With most Americans now opposing data centers in their communities, Paul argues the AI economy needs systemic reform of electricity tariffs, market structures, and regulatory policy — fast — before residential consumers are left holding the bill. Follow Paul on LinkedIn [https://www.linkedin.com/in/schusterpaul/].

Space-Based Data Centers: Understanding the Hype Behind SpaceX's IPO - Episode 124

Let us know how we're doing - text us feedback or thoughts on episode content [https://www.buzzsprout.com/2064554/fan_mail/new] SpaceX just went public at a $1.7 trillion valuation — and a big chunk of that bet is on orbital AI compute. Space-based data centers: GPUs in orbit, powered by unlimited solar, cooled by the vacuum of space, free from earthly permitting headaches. It sounds elegant. But does the business model actually work? In this episode, Paul breaks down the real promise and serious problems behind space-based data centers. He covers why terrestrial AI infrastructure is hitting hard limits on energy, water, land, and permitting — and why orbital compute is attracting serious capital as a result. Then he gets into the physics that SpaceX glosses over: the cooling problem. Not investment advice — just eight minutes of honest physics. Follow Paul on LinkedIn [https://www.linkedin.com/in/schusterpaul/].

The Hidden Carbon Cost of Enterprise AI: 5 Ways to Cut Emissions Without Cutting Usage - Episode 123

Let us know how we're doing - text us feedback or thoughts on episode content [https://www.buzzsprout.com/2064554/fan_mail/new] AI is quietly becoming one of the fastest-growing sources of corporate greenhouse gas emissions — yet not a single Fortune 500 sustainability report specifically calls it out. In this episode, Paul exposes the growing gap between AI's real emissions impact and how companies are (not) accounting for it. Drawing on a Capgemini Research survey of 2,000 executives — 48% of whom say AI has already materially impacted their corporate emissions, and 42% of whom have had to revisit their climate targets because of it — Paul explains why AI emissions are stuck in a measurement blind spot: an outdated GHG Protocol that buries AI usage, and near-total opacity from AI hyperscalers on model-specific carbon data. But there's good news: Paul lays out five practical ways enterprises can reduce their AI carbon footprint today — from right-sizing AI models and shifting workloads to cleaner grid regions, to temporal load shifting and prompt engineering audits — without reducing AI usage at all. Follow Paul on LinkedIn [https://www.linkedin.com/in/schusterpaul/].

Distributed Resources and the AI Data Center Build (Sachu Constantine - Vote Solar) - Episode 122

Let us know how we're doing - text us feedback or thoughts on episode content [https://www.buzzsprout.com/2064554/fan_mail/new] AI data centers are overwhelming the power grid — but the fastest, cheapest, and cleanest solution may already be sitting on rooftops. In this episode, Paul sits down with Sachu Constantine, Executive Director of Vote Solar, to explore why distributed solar and battery storage are the overlooked key to powering America's AI boom. Paul and Sachu unpack why utilities are struggling to keep pace with unprecedented data center load growth, why quick-fix solutions  fall short on cost, efficiency, and community impact, and how rooftop solar paired with storage at the grid edge can deliver speed, flexibility, and clean energy at scale. Want the full conversation? Catch the extended interview with Sachu Constantine on the sister podcast, More Than Eight Minutes. Follow Paul on LinkedIn [https://www.linkedin.com/in/schusterpaul/].