This seems to violate many of the OpenAI release guidelines (I am currently working on guiding a project through them, so am painfully aware of their stringency).

Notably it:

- Anthropomorphises the AI agent

- Claims 'actual' intelligence