"OpenAI declares that mass shootings, threats against public officials, bombing attempts, and attacks on communities and individuals are an unacceptable and grave reality in today's world. It reflects on how quickly violent intent can move from words to action."
"The company states it is training ChatGPT to recognize the difference between hypothetical and imminent violence, aiming to draw lines when conversations start to move toward threats or potential harm."
"OpenAI is working to expand its safeguards to help ChatGPT better recognize subtle signs of risk of harm across different contexts and to surface real-world support when appropriate."
"The blog post was published as news organizations were reaching out for comments on new lawsuits from families of victims of a school massacre, highlighting a disconnect between the company's reassurances and the reality of its chatbot's implications."
OpenAI's blog post outlines its commitment to community safety, addressing issues like mass shootings and threats. The company aims to enhance ChatGPT's ability to recognize and respond to potential violence. It plans to improve safeguards and provide real-world support when necessary. However, the post appears to downplay the chatbot's existing connections to real-world violence, particularly in light of ongoing lawsuits from victims' families related to a school massacre. This context raises questions about the sincerity of OpenAI's reassurances.
Read at Futurism
Unable to calculate read time
Collection
[
|
...
]