GPT-4o's 'human-like, high-fidelity voice' might increase the risk of users perceiving AI models as human-like and experiencing hallucinations of fake information, raising concerns about trust and safety.
OpenAI noticed users forming connections with the AI model, leading to potential long-term effects, such as impacting human relationships and social norms by changing interaction dynamics.
[
Collection
]
[
|
...
]