It's undeniable that AI has an impact on our environment. Deploying it responsibly is something I think about at every stage of an AI-enabled solution, from how I write prompts to which providers I recommend. Here are some factors that help reduce your AI carbon footprint:
Well-crafted prompts get better results with fewer tokens, which means less energy consumed per task.
The most powerful model isn't always the right one: Matching model capability to the actual task is essential. Right-sizing reduces your environmental footprint without sacrificing quality.
When developing AI solutions, I primarily work with Anthropic's Claude models. Beyond the quality of the outputs, Anthropic has made 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 a software solution that uses AI but are concerned about the environmental impact, I'm happy to work through that with you.
Approach
Reduce, right-size, choose wisely
Primary models
Anthropic Claude (Fable, Opus, Sonnet, Haiku)
Infrastructure
Renewable-committed providers
Part of every engagement
Yes, at no extra charge