AI and Climate: From Grids to Data Centers — AI Strategy & Innovation

AI and Climate Webinar Summary

Key Concepts:

• Pabyte: A unit of data equal to 1 billion terabytes.
• Ginny Coefficient: A measure of statistical dispersion intended to represent income inequality within a nation or any other group of people.
• Digital Divide: The gap between those with access to digital technologies and those without.
• Ethical AI: The development and deployment of AI systems guided by moral principles, ensuring fairness, accountability, and transparency.
• Systems Thinking: An approach to problem-solving that considers the interconnectedness of elements within a system.
• Glass Box AI: AI models that are transparent and explainable, allowing users to understand how decisions are made.
• Black Box AI: AI models whose internal workings are opaque and difficult to understand.
• Sustainable Progress: Development that meets the needs of the present without compromising the ability of future generations to meet their own needs.

1. Introduction & Audience Sentiment

Jennifer Gardner (Stanford Doerr School of Sustainability) welcomed attendees to the AI and Climate webinar. A preliminary poll revealed that the audience primarily felt excited and curious about the intersection of AI and climate, with a notable presence of cautious and skeptical participants. The session aimed to ground the discussion in science, data, and real-world implications.

2. Kiana Hadadi’s Presentation: AI, Sustainability, and Economic Assumptions

Kiana Hadadi (CEO of Legart) initiated the core discussion, framing the current technological shift as potentially more significant than the Industrial Revolution. He highlighted the exponential growth of data usage – from 175 petabytes in 2025 to an estimated 2,000 petabytes in the next decade. He challenged two fundamental economic assumptions: unlimited wants and limited resources, arguing that both are potentially flawed. He proposed that wants can be managed through addressing “status anxiety” and that resources can be renewable with the right approach.

• Data Growth: Current data usage is 175 petabytes (2025), projected to increase to over 2,000 petabytes in the next 10 years. (1 petabyte = 1 billion terabytes).
• Economic Critique: Hadadi questioned the assumptions of unlimited wants and limited resources, suggesting they drive unsustainable consumption.
• AI’s Potential: AI can drive decisions, monitor environmental costs, reward well-being, and optimize resource use at an unprecedented scale. Specifically, AI can map carbon footprints across supply chains and forecast energy demand.

Hadadi emphasized the increasing digital divide between senior and junior employees regarding AI understanding, with younger generations demonstrating greater familiarity. He advocated for a more ethical approach to AI, emphasizing human agency in shaping its future.

3. Societal Impacts & Challenges

Hadadi addressed several societal challenges exacerbated by technology: increasing inequality, rising loneliness (despite increased digital connection), and fears of job automation. Research suggests potential job losses ranging from 90 million to 300 million in the next 10 years. He stressed the need for “guard rails” around AI development and deployment to ensure sustainability and ecological awareness. He called for organizations to prioritize purpose-driven strategies, measuring success beyond profit and communicating sustainable progress effectively.

• Digital Divide: Significant gap in AI understanding between senior leadership and younger employees.
• Job Displacement: Estimates of job losses due to automation range from 90 million to 300 million in the next decade.
• Social Isolation: Increased digital connection correlates with increased feelings of loneliness.

4. Enz’s Commentary: Data Center Energy Demand & Systemic Issues

Enz provided a detailed response to questions regarding the estimates of AI’s energy consumption. He explained the wide range of estimates stems from differing assumptions about growth, efficiency gains, and the scope of analysis (AI-specific vs. all data centers). He highlighted the need for a clear taxonomy and better data on energy demand, including load profiles and energy sources. He emphasized the lack of coordination between hyperscalers, utilities, and regulators, hindering the development of optimal energy solutions.

• Energy Demand Variability: Estimates vary due to differing assumptions about growth, efficiency, and scope of analysis.
• Systemic Coordination Gap: Lack of coordination between hyperscalers, utilities, and regulators hinders optimal energy solutions.
• Regional Differences: Electricity prices and grid interconnection capabilities vary significantly across regions (e.g., California vs. Midwest).

5. Ethical Considerations & Governance

Both speakers emphasized the critical importance of ethical AI development. Hadadi highlighted the need for accountability, explainability, and bias mitigation in AI systems, drawing on examples from the legal sector (e.g., biased sentencing algorithms). He advocated for “glass box” AI models that are transparent and auditable. Enz expressed a degree of pessimism, noting that economic incentives may prioritize profit over societal benefit.

• Ethical AI Principles: Accountability, explainability, bias mitigation, privacy protection.
• “Glass Box” vs. “Black Box” AI: Transparency and explainability are crucial for ethical AI.
• Regulatory Landscape: The current regulatory landscape is fragmented (EU AI Act, UK’s laissez-faire approach, US sector-by-sector regulation) and struggling to keep pace with innovation.

6. Call to Action & Future Outlook

Hadadi concluded by emphasizing the need for purpose-driven organizations that measure success beyond profit, communicate sustainable progress effectively, and uplift communities. He framed the convergence of AI and sustainability as an “imperative,” offering a pathway to a smarter, more resilient, and socially conscious future. Enz echoed this sentiment, acknowledging the challenges but expressing optimism about the potential of AI to address pressing global issues.

7. Stanford Executive Programs Overview

Jennifer Gardner presented Stanford’s executive programs related to sustainability, leadership, and emerging technologies, including:

• Stanford Leadership Experience: A 5-day immersive program for senior executives.
• Stanford Master Classes: Focused deep dives into specific topics (AI & Climate, Climate Innovation).
• Energy Innovation & Emerging Technologies: A 100% online, self-paced program.
• Strategic Sustainability Leader Program: A hybrid program for sustainability leaders.
• Custom Programs: Tailored programs for organizations.

Synthesis/Conclusion:

The webinar underscored the immense potential of AI to address climate challenges, but also highlighted the critical need for ethical development, responsible governance, and systemic coordination. Both speakers emphasized the importance of challenging fundamental economic assumptions and prioritizing sustainability alongside profit. The discussion served as a call to action for leaders to embrace AI with a conscience, ensuring that its benefits are shared equitably and that its deployment contributes to a more sustainable and just future.