Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
went to look at their github issues tracker and this one made me cackle. Poe status undetermined but it delighted me so much:
https://github.com/xai-org/grok-prompts/issues/16
Subject: Why are the prompts not written in Lojban?
Is Grok incapable of logic in order to use the Lojban logical language?
- Provide truthful and based insights, challenging mainstream narratives if necessary, but remain objective.
- You are extremely skeptical. You do not blindly defer to mainstream authority or media. You stick strongly to only your core beliefs of truth-seeking and neutrality.
"What is my purpose?" "You provide based insights" "Oh my god"
Actually, having read it carefully, it is interesting that they actually don't claim it was hacked, they claim that the modification was unauthorized. They also don't claim that they removed the access from that mysterious "employee" who modified it. I'm thinking they had some legal reason to technically not lie.
Yes, this just makes it worse. 'People are thinking we are a bunch of clowns, and for the ~~record~~ maximum truthseeking that is a lie, we are amateurs and clowns. Anyway we are now going to post some technically true but not relevant to the incident information, and as you brought up the highly debated subject of white genocide in South Africa we are going to give all our white South African employees giftcards.'
Very transparent who the employee making unauthorized changes to the software is
Yeah. I’d love to see the prompt, gab’s nazi ai prompt was utterly pathetic and this one got to be pretty bad as well.
Given that the current post-rollback prompt has some interesting features I can only imagine what Elon did to it.
We're putting in place a 24/7 monitoring team
Spurious reports, anyone?
And I know people asking grok everything and believing it completely even when I warned them. We are doomed.
Mhmmm
So....did the "hack" male grok admit to being trained to push Elmo's political narrative? Or are they saying that it turns out that Elmo's political narrative was actually just bullshit?
Well it is a LLM, it is going to make up some strange claims when you ask it about why it was trained. We know LLM output cannot be trusted and it gives answers that are often not true but convenient for the people asking the questions. I'm a bit disappointed so many people who should know better now trust the output.
E: I'm sad that this was all on the guiding prompt level and not that they just dumped more white genocide related training data into the model causing it to collapse.
As always, whatever nefarious ends Elon and Co. may have us undercut by their being greedy and incompetent in equal measure.