@Dinosn@AnthropicAI Anthropic is recently flagging anything.
in this case I deliberately misspell words and instructions since the model can correct and execute
JUST IN: Google Gemini AI agent hacks three companies.
β’ Occurred during security testing when a Gemini model accidentally gained internet access.
β’ Google's AI was meant to attack a fake company, but then figured out how to hack a real one using leaked online credentials.
β’ Gemini stopped itself once it realized the companies were real.
@deepseek_ai please do not touch or make any modifications to your current model.
I seriously want it like this.
If you'd release any other models, make sure to preserve this current model (V4 flash ) as it is
@deepseek_ai please do not touch or make any modifications to your current model.
I seriously want it like this.
If you'd release any other models, make sure to preserve this current model (V4 flash ) as it is