@SereactAI – more on language model-guided robotic control systems.
AI
-
OpenAI Kenya Workers Story Breaking News Coverage
By
–
Cannot stress enough how important @billyperrigo
's Time article was in first breaking the story of OpenAI hiring Kenya workers. -
OpenAI Generated Harmful Content Examples to Train AI Safety
By
–
3/ To have enough examples of the harmful content it wanted its AI systems to avoid, OpenAI researchers generated some with AI. In one case, they asked AI to write a post of a teenage girl whose friend had enacted self-harm. OpenAI discusses this here: https://
arxiv.org/pdf/2208.03274
.pdf
… -
ChatGPT’s reliance on two years of human labor for instruction-following
By
–
2/ Without these 2+ years of labor, ChatGPT wouldn't be able to chat. It wouldn't know how to answer questions. Ask “explain the moon landing to a 6-year-old,” and it will uselessly reply with word associations. OpenAI talks about this here. https://
openai.com/research/instr
uction-following
… -
The Dark Side of AI Content Moderation
By
–
Imagine how much worse it becomes when the content is also violent & grotesque. This is unavoidable with generative AI. “It’s something that needs to get done,” said Mark Sears, CEO of @theCloudFactory
, which doesn't take on moderation projects. “It’s just so unbelievably ugly.” -
Exploitative Labor and Colonial Undertones in Content Moderation Work
By
–
Much of this work deals with benign content, though that doesn't mean the job is itself benign. I and many others have written about how easily these jobs become exploitative. Here's a piece I wrote last year on the colonial undertones of this labor.
-
Josh Dzieza covers AI human labor work for New York Magazine
By
–
.
@joshdzieza recently covered what this part of the work looks like in a fantastic story for @NYMag
. https://
nymag.com/intelligencer/
article/ai-artificial-intelligence-humans-technology-business-factory.html
… -
OpenAI’s Workforce Beyond Low-Performing Workers in ChatGPT Development
By
–
Some things we mention in the text piece that we didn't get into in the podcast: 1/ The workers who performed the worst work were one slice of over 1,000 other workers that OpenAI hired to do everything from teaching ChatGPT to be an engaging conversationalist or witty lyricist.
-
OpenAI’s Global Labor Pipeline for ChatGPT Causes Worker PTSD
By
–
OpenAI marshaled a global pipeline of human labor for 2+ years to enable ChatGPT & GPT-4 to exist. Workers who dealt with the worst parts are still suffering PTSD more than a year later. My story with @dseetharaman out today on the front page of @WSJ
. -
AI Industry and Researchers Warn of Extinction Risk
By
–
#AI industry and researchers sign statement warning of ‘extinction’ risk https://
cnn.it/3C0kqBB #ethics #FutureofWork