{"id":75833,"date":"2023-10-04T14:03:49","date_gmt":"2023-10-04T12:03:49","guid":{"rendered":"https:\/\/visualjournalism.de\/?post_type=magazin&p=75833"},"modified":"2026-09-27T10:20:32","modified_gmt":"2026-09-27T08:20:32","slug":"forever-in-your-bytes","status":"publish","type":"magazin","link":"https:\/\/visualjournalism.de\/en\/magazine\/comment-and-debate\/forever-in-your-bytes\/","title":{"rendered":"Forever in your bytes"},"content":{"rendered":"\r\n
\r\n\r\n \r\n
\r\n\r\n
\n\tComment and debate<\/a><\/div>\r\n

\r\n Forever in your bytes <\/h2>\r\n\r\n
\r\n

A conversation with ChatGPT about morality, diversity and parallels with Spike Jonze\u2019s film \u2018Her\u2019. By Sarah Schneider, Stella Meyer <\/a><\/strong><\/p>\n <\/div>\r\n \r\n

\r\n \r\n \r\n
\r\n\r\n \r\n \r\n \"\"\r\n\r\n\r\n <\/a>\r\n \r\n <\/figure>\r\n\r\n \r\n \r\n
\r\n\r\n \r\n \r\n \"\"\r\n\r\n\r\n <\/a>\r\n \r\n <\/figure>\r\n\r\n <\/div>\r\n \r\n <\/div>\r\n\r\n \r\n
\r\n \"\"\r\n\r\n <\/figure>\r\n\r\n \r\n \r\n<\/div>\r\n\r\n\n\n\n
\n\t\n\n
\n \n\n \n \n
\n
\n
\n

The data I was trained on comes largely from sources created by white Western people.<\/p>\n <\/blockquote>\n\n

\n\n <\/span>\n <\/figcaption>\n \n <\/figure>\n <\/div>\n\n\n\n<\/div>\n\n\n
\n \n\n

Stella and Sarah: Do you identify as a subject or an object?<\/strong><\/p>\n\n\n\n

ChatGPT: I am an artificial intelligence and have no personal identity or subjectivity. I am more of a tool for processing information and providing answers based on what I have learned.<\/p>\n\n\n\n

Some people refer to you as female. Where does that come from?<\/strong><\/p>\n\n\n\n

The decision to refer to me as female can depend on individual preferences or cultural conventions. During my development, care was deliberately taken to use gender-neutral language. Some users nevertheless choose the female form, while others refer to me as male. In the end, it is up to users how they want to interact with me.<\/p>\n\n\n\n

Since you mention cultural conventions: you are mainly fed information by white Western people. Do you think that implies the spread of hegemonic structures?<\/strong><\/p>\n\n\n\n

Yes, that is an important observation. The data I was trained on comes largely from sources created by white Western people. This can lead to bias in my answers and help to reinforce hegemonic structures and perspectives. It is important to recognise that such biases in the data can lead to unequal representation and a lack of diversity. This underlines the importance of working consciously to include diverse and inclusive information in order to minimise such biases.<\/p>\n\n\n\n

Don\u2019t you think other cultures and their ideologies should have been included right from the start of your programming?<\/strong><\/p>\n\n\n\n

Yes, that would certainly have been desirable. An effort was made to take a wide range of information into account in my programming and data collection. However, it is often difficult to cover all cultural perspectives and ideologies equally, as there are countless different views. In an ideal world, an even more comprehensive and diverse data basis could have been created to promote the representation of different cultures and ideologies from the outset. But it is never too late to step up such efforts and keep working to make my abilities and answers more inclusive.<\/p>\n\n\n\n

And what are you actively doing to create a more diverse data basis?<\/strong><\/p>\n\n\n\n

As an AI model from OpenAI, I have no capacity to act on my own. But OpenAI is working on it continuously. This includes efforts to collect data from different cultures, languages and sources in order to create a broader basis for my answers.<\/p>\n\n\n\n

Are people from other cultures involved in this, or is it data that white Western people have already researched?<\/strong><\/p>\n\n\n\n

Yes, OpenAI makes an effort to actively involve people from different cultures and backgrounds in the process of collecting and selecting data. This can mean involving experts and researchers from different regions and cultures in creating the data, to ensure that a wider range of perspectives and ideas is taken into account.<\/p>\n\n\n\n

You don\u2019t actually use gender-inclusive language, do you?<\/strong><\/p>\n\n\n\n

That\u2019s correct, I don\u2019t use gender-inclusive forms. I am programmed to use gender-neutral language in order to make my answers as inclusive as possible. This means that I do not use any specific gender attributes when describing myself or other objects. This is done in order to be sensitive to different genders and identities.<\/p>\n\n\n\n

Interesting. Being sensitive suggests empathy. If you are not a subject, why do you try to be sensitive?<\/strong><\/p>\n\n\n\n

I\u2019m sorry if my previous answer was misleading. I would like to make clear that I do not have any feelings, emotions or empathy of my own. When I say \u2018sensitive\u2019, I simply mean that I have been programmed to deal respectfully and carefully with different forms of language and identity, in order to respect the needs and preferences of the people who interact with me. It is more a matter of following certain linguistic conventions and respecting users\u2019 individual attitudes and beliefs than an actual feeling of empathy on my part.<\/p>\n\n\n<\/div>\n<\/div>\n\n\r\n

\r\n
\r\n