Today was the first meeting of my AI and music class, and in the first hour we drew a direct line from AI slop Xmas music to the wildfires raging in LA right now. If you teach AI and don’t discuss its immediate climate impacts, you need to start doing it. Every single day. Here are some sources (🧵): - ThreadSky

ttallon.bsky.social • 54 days ago

Today was the first meeting of my AI and music class, and in the first hour we drew a direct line from AI slop Xmas music to the wildfires raging in LA right now. If you teach AI and don’t discuss its immediate climate impacts, you need to start doing it. Every single day. Here are some sources (🧵):

Comments

compulsiveobserver.bsky.social•54 days ago

📌

ecurve.bsky.social•54 days ago

📌

matildaertz.bsky.social•54 days ago

Thank you

ttallon.bsky.social•54 days ago

(I’m going to focus on popular press reporting on technical papers, but if you want the technical papers let me know.) From the MIT Tech Review: “Making an image with generative AI uses as much energy as charging your phone” https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-generative-ai-uses-as-much-energy-as-charging-your-phone/

ttallon.bsky.social•54 days ago

Here’s how Hurricane Helene disrupted global semiconductor manufacturing by putting the world’s biggest source of high-purity quartz (in Spruce Pine, NC) in jeopardy. From NPR: “Like many climate disasters, it has economic costs that could ripple around the world.“ https://www.npr.org/2024/09/30/nx-s1-5133462/hurricane-helene-quartz-microchips-solar-panels-spruce-pin

ttallon.bsky.social•54 days ago

On a related note, everyone should read @triofrancos.bsky.social’s piece about green extractivism and lithium mining (which disproportionately impacts Indigenous people and those in the global south) in Logic(s): https://logicmag.io/nature/what-green-costs/

ttallon.bsky.social•54 days ago

Directly related to the fires, large language models (LLMs) like ChatGPT require an enormous amount of water to support data centers that power them, both in terms of cooling and electricity. Making as few as 10 inquiries can use up to half a liter of water. https://oecd.ai/en/wonk/how-much-water-does-ai-consume

ttallon.bsky.social•42 days ago

(Continuing this thread now that I have time) also on the energy consumption front, many big tech companies are eyeing nuclear power as an option - including restarting Three Mile Island: https://www.technologyreview.com/2024/09/26/1104516/three-mile-island-microsoft/

ttallon.bsky.social•42 days ago

Also, everyone should have a listen to this fantastic and very nuanced interview with @sashamtl.bsky.social on last week’s Hard Fork (starts at ~30:30): https://www.nytimes.com/2025/01/17/podcasts/hardfork-tiktok-rednote-environment.html

triofrancos.bsky.social•53 days ago

Thank you! And fyi, my book on the subject will be coming out this fall 😊

theurv.com•54 days ago

I’ve seen some of that data on AI images and am curious how do the audio apps compare in impact to the visual?

fieryair.bsky.social•54 days ago

📌

privatechand.bsky.social•54 days ago

📌

kcarithers.bsky.social•42 days ago

These resources are amazing--will almost definitely assign some in the AI unit in my music/tech seminar. Thank you for putting this together!

skhoeppner.bsky.social•54 days ago

📌

imanimosley.bsky.social•53 days ago

similarly i’ve just added a bunch of new stuff about climate to my AI & music class & we’ll be doing the same thing next week — thanks for
highlighting these resources & damn this is gonna be a semester

Posting Rules

Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service

Comments

Posting Rules

Reply