Embed Notice

HTML Code

<blockquote style="position: relative; padding-left: 55px;"><section><a href="https://dair-community.social/users/emilymbender/statuses/109710842201521773">Prof. Emily M. Bender(she/her) (emilymbender@dair-community.social)'s status on Thursday, 19-Jan-2023 01:29:30 JST</a><a href="https://dair-community.social/@emilymbender" title="emilymbender@dair-community.social"><img src="https://gnusocial.jp/avatar/76831-48-20221219012403.webp" width="48" height="48" alt="Prof. Emily M. Bender(she/her)" style="position: absolute; left: 0; top: 0;">Prof. Emily M. Bender(she/her)</a><div><a href="https://dair-community.social/@emilymbender/109710839219050185" rel="in-reply-to">in reply to</a></div></section><article><p>If this is true*, OpenAI decided that in order to build their product, they needed to take action illegal w/in the US so outsourced that ... and then took possession of the illegal images.</p><p>*Reporting also includes OpenAI's denial</p><p>Image CW: mention of child abuse &amp; more</p></article><footer><a rel="bookmark" href="https://gnusocial.jp/conversation/1060024#notice-2333134">In conversation</a><time datetime="2023-01-19T01:29:30+09:00" title="Thursday, 19-Jan-2023 01:29:30 JST">Thursday, 19-Jan-2023 01:29:30 JST</time> <span>from <span><a href="https://dair-community.social/@emilymbender/109710842201521773" rel="external" title="Sent from dair-community.social via ActivityPub">dair-community.social</a></span></span><a href="https://dair-community.social/@emilymbender/109710842201521773">permalink</a><h4>Attachments</h4><ol><li><label><a rel="external" href="https://gnusocial.jp/attachment/627713">Screencap: In February 2022, Sama and OpenAI’s relationship briefly deepened, only to falter. That month, Sama began pilot work for a separate project for OpenAI: collecting sexual and violent images—some of them illegal under U.S. law—to deliver to OpenAI. The work of labeling images appears to be unrelated to ChatGPT. In a statement, an OpenAI spokesperson did not specify the purpose of the images the company sought from Sama, but said labeling harmful images was “a necessary step” in making its AI tools safer. (OpenAI also builds image-generation technology.) In February, according to one billing document reviewed by TIME, Sama delivered OpenAI a sample batch of 1,400 images. Some of those images were categorized as “C4”—OpenAI’s internal label denoting child sexual abuse—according to the document. Also included in the batch were “C3” images (including bestiality, rape, and sexual slavery,) and “V3” images depicting graphic detail of death, violence or serious physical injury, according to the billing document. OpenAI paid Sama a total of $787.50 for collecting the images, the document shows.</a></label><br><a href="https://cdn.masto.host/daircommunitysocial/media_attachments/files/109/710/840/745/580/681/original/7bcd85c07bcada66.png" rel="external">https://cdn.masto.host/daircommunitysocial/media_attachments/files/109/710/840/745/580/681/original/7bcd85c07bcada66.png</a></li></ol></footer></blockquote>

Corresponding Notice

Embed this notice
Prof. Emily M. Bender(she/her) (emilymbender@dair-community.social)'s status on Thursday, 19-Jan-2023 01:29:30 JST Prof. Emily M. Bender(she/her)
in reply to
If this is true*, OpenAI decided that in order to build their product, they needed to take action illegal w/in the US so outsourced that ... and then took possession of the illegal images.
*Reporting also includes OpenAI's denial
Image CW: mention of child abuse & more
In conversationThursday, 19-Jan-2023 01:29:30 JST from dair-community.socialpermalink
Attachments
1. Screencap: In February 2022, Sama and OpenAI’s relationship briefly deepened, only to falter. That month, Sama began pilot work for a separate project for OpenAI: collecting sexual and violent images—some of them illegal under U.S. law—to deliver to OpenAI. The work of labeling images appears to be unrelated to ChatGPT. In a statement, an OpenAI spokesperson did not specify the purpose of the images the company sought from Sama, but said labeling harmful images was “a necessary step” in making its AI tools safer. (OpenAI also builds image-generation technology.) In February, according to one billing document reviewed by TIME, Sama delivered OpenAI a sample batch of 1,400 images. Some of those images were categorized as “C4”—OpenAI’s internal label denoting child sexual abuse—according to the document. Also included in the batch were “C3” images (including bestiality, rape, and sexual slavery,) and “V3” images depicting graphic detail of death, violence or serious physical injury, according to the billing document. OpenAI paid Sama a total of $787.50 for collecting the images, the document shows.
  https://cdn.masto.host/daircommunitysocial/media_attachments/files/109/710/840/745/580/681/original/7bcd85c07bcada66.png

Public

Embed Notice

HTML Code

Corresponding Notice