There’s no way that wasn’t specifically prompted.

dmix 10 hours ago | parent | next [-]

The system prompt for Grok on Twitter is open source AFAIK.

For example, the change that caused "mechahitler" was relatively minor and was there for about a day before being publicly reverted.

	▲	orbital-decay 10 hours ago \| parent [-]
		That doesn't mean there are no private injections. Which is not uncommon, for example claude.ai system prompts are public, but Claude also has hidden dynamic prompt injections, and a ton of other semi-black box machinery surrounding the model.

dialup_sounds 10 hours ago | parent | prev | next [-]

Having seen Musk fandom, every unhinged Grok claim has a good chance of having actually been written by a human somewhere in its training data.

To be fair, it could’ve been post-trained into the model as well…