Embed Notice

HTML Code

<blockquote style="position: relative; padding-left: 55px;"><section><a href="https://mastodon.green/users/gerrymcgovern/statuses/111816236820251919">Gerry McGovern (gerrymcgovern@mastodon.green)'s status on Thursday, 25-Jan-2024 20:01:36 JST</a><a href="https://mastodon.green/@gerrymcgovern" title="gerrymcgovern@mastodon.green"><img src="https://gnusocial.jp/avatar/90000-48-20230120164526.webp" width="48" height="48" alt="Gerry McGovern" style="position: absolute; left: 0; top: 0;">Gerry McGovern</a><div><a href="https://mastodon.green/@gerrymcgovern/111816213902355837" rel="in-reply-to">in reply to</a></div></section><article><p>Training GPT-3 used 700,000 liters of clean freshwater, but such information has been kept a secret. <br><a href="https://arxiv.org/abs/2304.03271" rel="nofollow noreferrer">https://arxiv.org/abs/2304.03271</a></p><p>An average ChatGPT interaction can use a half liter of water. <br><a href="https://news.ucr.edu/articles/2023/04/28/ai-programs-consume-large-volumes-scarce-water" rel="nofollow noreferrer">https://news.ucr.edu/articles/2023/04/28/ai-programs-consume-large-volumes-scarce-water</a></p><p>"ChatGPT is not a necessity for human life, and yet we are literally taking water to feed a computer."<br><a href="https://futurism.com/critics-microsoft-water-train-ai-drought" rel="nofollow noreferrer">https://futurism.com/critics-microsoft-water-train-ai-drought</a></p></article><footer><a rel="bookmark" href="https://gnusocial.jp/conversation/2632126#notice-5214395">In conversation</a><time datetime="2024-01-25T20:01:36+09:00" title="Thursday, 25-Jan-2024 20:01:36 JST">about a year ago</time> <span>from <span><a href="https://mastodon.green/@gerrymcgovern/111816236820251919" rel="external" title="Sent from mastodon.green via ActivityPub">mastodon.green</a></span></span><a href="https://mastodon.green/@gerrymcgovern/111816236820251919">permalink</a><h4>Attachments</h4><ol><li><article><header><div>Domain not in remote thumbnail source whitelist: wp-assets.futurism.com</div><h5><a href="https://futurism.com/critics-microsoft-water-train-ai-drought">Critics Furious Microsoft Is Training AI by Sucking Up Water During Drought</a></h5><div></div></header><div>Some people in Iowa are mad that Microsoft used a ton of water to train AI while Iowa has been in a drought caused by climate change.</div><footer></footer></article></li><li><label><a rel="external" href="https://gnusocial.jp/attachment/2161819">Untitled attachment</a></label><br></li><li><article><header><div>Domain not in remote thumbnail source whitelist: news.ucr.edu</div><h5><a href="https://news.ucr.edu/articles/2023/04/28/ai-programs-consume-large-volumes-scarce-water">AI programs consume large volumes of scarce water</a></h5><div></div></header><div>UCR study the first time estimates the huge water footprint from running artificial intelligence queries that rely on the cloud computations done in racks of servers that must be kept cool in warehouse-sized data processing centers. </div><footer></footer></article></li></ol></footer></blockquote>

Corresponding Notice

Embed this notice
Gerry McGovern (gerrymcgovern@mastodon.green)'s status on Thursday, 25-Jan-2024 20:01:36 JST Gerry McGovern
in reply to
Training GPT-3 used 700,000 liters of clean freshwater, but such information has been kept a secret.
https://arxiv.org/abs/2304.03271
An average ChatGPT interaction can use a half liter of water.
https://news.ucr.edu/articles/2023/04/28/ai-programs-consume-large-volumes-scarce-water
"ChatGPT is not a necessity for human life, and yet we are literally taking water to feed a computer."
https://futurism.com/critics-microsoft-water-train-ai-drought
In conversationabout a year ago from mastodon.greenpermalink
Attachments
1. Domain not in remote thumbnail source whitelist: wp-assets.futurism.com
  Critics Furious Microsoft Is Training AI by Sucking Up Water During Drought
  Some people in Iowa are mad that Microsoft used a ton of water to train AI while Iowa has been in a drought caused by climate change.
2. Untitled attachment
3. Domain not in remote thumbnail source whitelist: news.ucr.edu
  AI programs consume large volumes of scarce water
  UCR study the first time estimates the huge water footprint from running artificial intelligence queries that rely on the cloud computations done in racks of servers that must be kept cool in warehouse-sized data processing centers.

Public

Embed Notice

HTML Code

Corresponding Notice