There's a serious proposal along the same lines: https://www.unicode.org/L2/L2025/25241-ai-watermarks.pdf

I feel like there is an unofficial version of AGTI already in place for certain AI providers.

Whenever I generate a large amount of code, there is a ~20% chance that my editor will pop a warning "Some unicode characters in this file could not be saved in the current codepage".

I suggest taking a look at the raw outputs of a major AI provider in a hex editor. That (zero-width) whitespace could be hiding a lot of information.

▲

marginalia_nu 6 hours ago | parent | prev | next [-]

(Flashbacks from the horrors in the Byte Order Mark wars)

	▲	xnorswap 6 hours ago \| parent \| next [-]
		That still trips us up regularly to this day.
	▲	dionian 6 hours ago \| parent \| prev [-]
		Oh god

▲

lxgr 6 hours ago | parent | prev | next [-]

Maybe considered serious by its proponents.

▲

wizzwizz4 2 hours ago | parent | prev [-]

> For example, every other letter in this sentence is U+2060.

No it isn't! The PDF renderer has stripped them out.