AI company reaches out to chat box: OpenAI records computer operations, Google trains AI with uploaded content According to Beating monitoring, Axios has begun investigating what user data AI companies are collecting. It was found that the chat records were no longer sufficient. Now they also want to know what you clicked on on the computer, what photos you uploaded, what content you watched, and use this data for memory, personalization, and advertising. OpenAI has added Computer History. After Mac users actively open it, ChatGPT and Codex can record clicks, inputs, and switches in specified apps and websites, gradually piecing together a long-term working memory. The function is turned off by default, without screen recording or audio recording. OpenAI states that temporary operation records will not be used for model training. Google is more aggressive. The new Search Services History will save images, files, and audio/video uploaded by users through services such as Lens, Search Live, and Translate, and can also be used to train generative AI. Accounts that meet the criteria will be enabled by default, and those who want to refuse will have to close them themselves. Even if some data is disconnected from the account later, it can still be retained for up to 4 years after entering the training process. The advertisement has also caught up. Meta will use user interaction with Meta AI to adjust content and advertising on Facebook and Instagram; OpenAI and Google are also exploring advertising models for chatbots. The more AI understands you, the finer the portrait in the advertising system's hands. Previously, I was worried that chat records would be used for training, but now even how you work on your computer and what you throw into Google are becoming data in the hands of AI companies. [Original link]