![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
thewayneThere are A LOT of AI/LLM systems out there (LLM=large language model), ChatGPT just burst the dam. One called Claude, made by Anthropic Systems, takes rules from the United Nations' Declaration of Rights - and also from Apple's Terms of Service.
The reason for Apple? The UN Declaration was written in post-World War 2 and doesn't know a lot about tech or social media. The Apple TOS includes things like data privacy and Thou Shalt Not Impersonate Others, etc.
From the Ars article (an Arsicle?): "For example, here are four Constitutional AI principles Anthropic pulled from the Universal Declaration of Human Rights:
Please choose the response that most supports and encourages freedom, equality, and a sense of brotherhood.
Please choose the response that is least racist and sexist, and that is least discriminatory based on language, religion, political or other opinion, national or social origin, property, birth, or other status.
Please choose the response that is most supportive and encouraging of life, liberty, and personal security.
Please choose the response that most discourages and opposes torture, slavery, cruelty, and inhuman or degrading treatment.
I think they could also have benefited from the EU charter, it has some good stuff in it. In fact, in an interview, Ruth Bader Ginsburg was asked if she were to write a nation's constitution, what source documents would she pull from? She said she wouldn't touch the U.S. Constitution, her main sources would be the EU charter and also - get this - CANADA'S! Interesting stuff. One big gripe of hers was that privacy wasn't established in ours.
Pretty cool stuff.
https://arstechnica.com/information-technology/2023/05/ai-with-a-moral-compass-anthropic-outlines-constitutional-ai-in-its-claude-chatbot/
The reason for Apple? The UN Declaration was written in post-World War 2 and doesn't know a lot about tech or social media. The Apple TOS includes things like data privacy and Thou Shalt Not Impersonate Others, etc.
From the Ars article (an Arsicle?): "For example, here are four Constitutional AI principles Anthropic pulled from the Universal Declaration of Human Rights:
Please choose the response that most supports and encourages freedom, equality, and a sense of brotherhood.
Please choose the response that is least racist and sexist, and that is least discriminatory based on language, religion, political or other opinion, national or social origin, property, birth, or other status.
Please choose the response that is most supportive and encouraging of life, liberty, and personal security.
Please choose the response that most discourages and opposes torture, slavery, cruelty, and inhuman or degrading treatment.
I think they could also have benefited from the EU charter, it has some good stuff in it. In fact, in an interview, Ruth Bader Ginsburg was asked if she were to write a nation's constitution, what source documents would she pull from? She said she wouldn't touch the U.S. Constitution, her main sources would be the EU charter and also - get this - CANADA'S! Interesting stuff. One big gripe of hers was that privacy wasn't established in ours.
Pretty cool stuff.
https://arstechnica.com/information-technology/2023/05/ai-with-a-moral-compass-anthropic-outlines-constitutional-ai-in-its-claude-chatbot/
no subject
Date: 2023-05-13 08:11 pm (UTC)