Opening the Black Box: Quantifying Closed-Model AI Emissions with CLEER
AI use is booming, but companies are adopting AI faster than they can account for its planetary impacts.
The most popular models that are used every day publish almost nothing about their energy use or carbon footprint. As a result, sustainability teams end up estimating emissions based on their spending…or leaving them out of the inventory entirely.
In this 60-minute webinar, Sustainable AI Group (SAIG) and Etsy show what it looks like to measure those emissions in a scientific way and act on them in an enterprise setting.
Sasha Luccioni, SAIG’s Co-founder and Chief Scientific Officer, will kick things off with a broader perspective about where AI sustainability stands today and why it’s so hard to estimate the impacts of closed models. Sasha and Nidhal Jegham, SAIG’s Research Engineer, will then introduce CLEER (Closed-model Latent Energy Estimation Range), SAIG's new approach. Together, they’ll explain what makes CLEER credible and how organizations can adopt it, from disclosing factors to mapping onto their own AI usage data.
Etsy's Sam Brundrett (Senior Manager, Environmental Lead) and Emily Sommer (Staff Systems Architect in Sustainable Computing) will then share how Etsy approaches AI's environmental impact: where AI sits in its sustainability strategy, how it is using CLEER in reporting and internal decisions, and an engineering example of reducing AI's footprint.
The session will end with an extended live Q&A with all speakers, moderated by Boris Gamazaychikov, SAIG’s Co-founder and CEO.
What you’ll take away
Why emissions from closed AI models have been so hard to measure, and how CLEER estimates them
How to use CLEER in sustainability reporting and in day-to-day decisions such as model choice
How Etsy puts model-level emissions next to cost, and what its engineers do with that data
A real example of an engineering change that reduced AI's impact
Who should attend
Sustainability and ESG reporting teams who need a defensible number for AI emissions
Engineering, platform and GreenOps practitioners looking for ways to cut AI's footprint without guesswork
Speakers
Dr. Sasha Luccioni, Co-Founder & Chief Scientific Officer, SAIG
Nidhal Jegham, Research Engineer, SAIG
Sam Brundrett, Senior Manager, Environmental Lead, Etsy
Emily Sommer, Staff Systems Architect in Sustainable Computing, Etsy
Moderator
Boris Gamazaychikov, Co-Founder & CEO, SAIG