It's undeniable that AI has an impact on our environment. That's not a reason to avoid it — but it is a reason to be thoughtful. Deploying AI responsibly is something I think about at every stage of a project, from how I write prompts to which providers I recommend.
Well-crafted prompts get better results with fewer tokens — which means less energy consumed per task. I design prompts that are specific, structured, and purposeful. Clear, deliberate instructions beat trial-and-error iteration every time, and not just for quality reasons.
The most powerful model isn't always the right one. Matching model capability to the actual task is something I factor into every solution I design. A smaller, faster model handling a straightforward classification task doesn't need to be swapped out for a frontier model just because one exists. Right-sizing reduces costs and environmental footprint without sacrificing quality.
I primarily work with Anthropic's Claude models. Beyond the quality of the outputs, Anthropic has made meaningful commitments around responsible AI development and infrastructure sustainability. That matters to me when I'm recommending tools to clients.
Part of every engagement
If you're thinking about building an AI solution but are concerned about the environmental impact, I'm happy to work through that with you — it's part of any project, not a separate line item.
Approach
Reduce, right-size, choose wisely
Primary models
Anthropic Claude (Opus, Sonnet, Haiku)
Infrastructure
Renewable-committed providers
Part of every engagement
Yes — no extra charge