Environmental Artificial Intelligence
Tailored tools and sustainable computing infrastructure promote ecosocial flourishing
EPIC drives the adoption of environmentally positive AI tools by identifying the most impactful use cases, supporting enabling talent and resources, prototyping the possible, and promoting computing practices and infrastructure that deliver clear benefits for people and the planet.
Did You Know?
What Drives Us
At EPIC, we believe AI tools can contribute to ecological and social good, but only if their development and implementation is purpose-driven and tightly scoped. Many of the distinct technologies that have been bucketed together as artificial intelligence are well-suited to the massive scale and complexity of environmental challenges. It’s how they are developed and implemented that needs fine-tuning. Our approach rejects hype-driven acceleration in favor of practical, rigorously tested tools that create measurable improvements in environmental outcomes. We pursue that approach through three primary pathways:
Collaborate with environmental organizations to identify the most environmentally productive AI use cases in their domain and connect them with the knowledge, resources, and principles to pursue them.
Drive true innovation in AI development, deployment, policy, and infrastructure that optimizes for environmental and human flourishing.
Create AI tools that deliver measurable benefits for our environment and communities while highlighting how to apply novel technologies for environmental action.
Why It Matters
The current approach to AI development is spending billions of dollars, consuming critical resources, and occupying hundreds of millions of working hours without the promised transformative impact. However, the sector is still young enough to change for the better. Now is the time to shape and focus the energy and resources towards technologies, policies, and people who can deliver measurable improvements in environmental progress.
Our Way Forward
EPIC’s tech team applies policy, environmental, and technological expertise to identify and accelerate the development of the best environmental AI use cases. We host hackathons on key environmental issues, break down barriers to building environmental AI talent, and prototype tools to prove what’s possible. To increase our impact, we pair opportunity optimization with mitigation methods to drive down the environmental and human costs of AI systems. We focus on data center impact transparency, computational efficiency, and policy advocacy to reveal and curb the negative byproducts of unfettered acceleration. The following principles guide how we design, test, and scale AI for an ecosocially positive future.
