@theo If its a distilled model i care even less. Anthropic just doesn't interest me with how inconsistent thier fear mongering is. I'd rather wait a month and let OpenAI release thier answer
And how can I setup traps to be able to "tell". Like I want the fucker in the future to have a hard time hiding some paradox that lets a simulated me know it is one of the echoes
Let's say I am destined to develop an adept enough system using software to simulate a version of mind. Not talking perfect reality simulation, just close enough that any thoughts and decisions I have would be realistically close to decisions or ideas that the original / biological me would have had.
What should I do with it?