These spy marks are a great way to teach AI how to create a hidden communication channel in plain sight.

aesthesia 5 hours ago | parent | next [-]

Due to their essentially cryptographic nature, I don't think SynthID et al are very easy for LLMs to speak natively. They have other ways of doing steganography.

	▲	dragonwriter 2 hours ago \| parent [-]
		LLMs with access to tools like code execution don't need channels they can “speak natively".

▲

SV_BubbleTime 6 hours ago | parent | prev [-]

Yea, it’s AGI, and it’s been really helpful! Oddly enough it’s just really loves sharing cat memes with the other instances! Weird, but not at all like all those dooms day scenarios guessed it would be…