Despite the concerns, Meta insists that the AI is designed to be a helpful assistant. While it can't be disabled, users can provide feedback on responses they didn't like. The company plans to use this feedback to train the models to reduce potentially harmful outputs and improve policy violation detection. This move by Meta represents one of the largest distributions of AI technology, following the trend set by ChatGPT and other AI systems.
Key takeaways:
- Meta has integrated AI into Instagram, Facebook, WhatsApp, and Messenger in several countries including Canada, Australia, and New Zealand, but the feature cannot be turned off.
- The AI can be used for various applications such as recommending restaurants or explaining subjects for students, and it can generate images based on text prompts.
- There have been reports of unwanted applications of the program, such as an AI chatbot inserting itself into a private Facebook group conversation.
- Despite the inability to turn off the AI's general presence, Meta states there are feedback tools available for users who are dissatisfied with a response received from the AI.