Event JSON
{
"id": "b17a9fba6a2ed5c8de0680f58e00509ffed368141500b3befc11fd4664a3931b",
"pubkey": "424c46da23fd8ddcd5486ee393bc5f9f9b551d6af14fefb0f928f04908c75771",
"created_at": 1775562571,
"kind": 1,
"tags": [
[
"proxy",
"https://lipn.info/users/cazencott/statuses/116363268693853313",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.ditto.pub"
]
],
"content": "Hey hive mind, do we have any numbers on training data size / model size / training and query energy consumption of the most recent \"AI\" tools (the ones sold by Anthropic, OpenAI and the like)?\n https://www.technologyreview.com/2025/05/20/1116327/ai-energy-usage-climate-footprint-big-tech/ is almost a year old now, same for most evaluations on https://huggingface.co/spaces/AIEnergyScore/Leaderboard\n\nI've found https://www.simonpcouch.com/blog/2026-01-20-cc-impact/ for Claude Code but that's a lot of guesswork",
"sig": "913fd0951436e3acf7404f0db9a1e23cafcad79cc66f5e7af354baafe699453c79c7c3afc23bcc63f9dc38f1b57e8d64d27b719065b700dd701c411241e49332"
}