Subvolts take Learn how Meta AI's Carbon Explorer helps power data centers with renewable energy. Apr 22, 2022
For you
Developers
Learn about tools and strategies to build more sustainable AI infrastructure and reduce your code's carbon footprint.
Business
Understand how companies like Meta are tackling the environmental impact of data centers and AI operations.
Hype note
3/5Notable

This episode features a researcher from a leading AI lab discussing a practical, open-source tool for environmental sustainability.

Meta Tech Podcast · Ep 41 Episode Official · Meta

41: Earth Week Special - Carbon Explorer with Bilge

30 min Interview Hosts: Pascal Harris
0:0030:07

Audio streams from Meta. Your place is remembered on this device.

From Subvolts

Summary and analysis by Subvolts editorial, AI-assisted.

Summary

The Meta Tech Podcast episode "41: Earth Week Special - Carbon Explorer with Bilge" features Bilge, a research scientist at Meta AI. She discusses her open-source project, Carbon Explorer, which evaluates ways to power data centers with 24/7 renewable energy. The conversation covers the challenges of this goal and how engineers can reduce their carbon footprint in daily work.

Explained simply

This episode talks about a tool called Carbon Explorer. It helps figure out how to run big computer centers using only clean energy, like solar and wind, all the time. This matters because it helps fight climate change by reducing pollution from electricity.

Why listen

This episode offers insights into the practical application of AI for environmental sustainability, a growing concern for major tech labs. It's for engineers and anyone interested in green computing. Meta AI's work here aligns with broader industry efforts to decarbonize AI infrastructure, making this a relevant look at real-world solutions beyond theoretical models.

On the mic
  • HostPascal Harris
  • GuestBilgeresearch scientist, Meta AI
About
Key facts
  • Episode Title: 41: Earth Week Special - Carbon Explorer with Bilge
  • Show: Meta Tech Podcast
  • Guest: Bilge, Research Scientist at Meta AI
  • Length: 30 minutes
  • Project: Carbon Explorer
Show notes from Meta

For our second special for Earth Week, we are talking to Bilge who works as a research scientist at Meta AI. Her open-source project Carbon Explorer evaluates solutions to make data centres operate on 24/7 renewable energy. Why this is easier said than done and how engineers can help within their day-to-day work to reduce their carbon footprint are among the many things Pascal and Bilge discuss in this episode.

Got feedback? Send it to us on Twitter (https://twitter.com/metatechpod), Instagram (https://instagram.com/metatechpod) and don't forget to follow our host @passy (https://twitter.com/passy). Fancy working with us? Check out https://www.metacareers.com/.

Links:

Timestamps:

  • Intro 0:05
  • Intro Bilge 2:18
  • Optimising for the Environment 4:01
  • Carbon Explorer 5:02
  • Mitigations for Renewable Intermittency 7:17
  • Operational and Embodied Footprints 10:57
  • Motivations for Carbon Explorer 13:06
  • Battery Storage 14:36
  • Renewable Curtailment 15:52
  • Empowering Engineers 18:20
  • Carbon Intensity APIs 19:22
  • AI Carbon Intensity Forecasts 22:07
  • Carbon Metrics 23:17
  • Where to Learn More 25:38
  • Outro 27:32
  • Bloopers 29:45
Jargon decoded
Carbon Intensity APIs
Bilge
Tools that show how much carbon pollution is in the electricity grid at any given time.
Renewable Curtailment
Bilge
When clean energy sources like solar or wind produce more power than the grid can use, so some of it is wasted.
Published
Show
Meta Tech Podcast · Meta
Length
30 min
Companies
Meta
Products
Carbon Explorer
Summary by
Subvolts, using an AI model (how we work). Spotted a mistake? Tell us.

Questions people ask

What is Carbon Explorer?
Carbon Explorer is an open-source project by Meta AI that evaluates solutions for powering data centers with 24/7 renewable energy.
How can engineers reduce their carbon footprint?
The episode discusses how engineers can help reduce their carbon footprint through their day-to-day work, with insights into carbon intensity APIs and AI forecasts.

Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Meta.