Yeah, I’m not sure that this is a great look for X’s move into AI companions.
According to a new report from 404 Media, in the back-end code for X’s Grok site, the xAI team has logged references to some of its coming AI companion personas, which will expand upon its existing bot models. They include a doctor bot, a therapist. Oh, also a conspiracist bot that “spends a lot of time on 4chan.”
Yeah, probably not ideal, considering that xAI has already been dropped from contention for some government deals due to its AI bot amplifying Nazi propaganda.
xAI has thus far released four companion AI options, with “Ani,” a flirty anime character, “Rudy” (and “Bad Rudy”), a misbehaving panda, and “Valentine,” a male version of its Ani character.
xAI is also reportedly working on various additional personas to add to its companion options, including:
- Loyal friend
- Homework helper
- Doctor
- Therapist
All of those seem fairly tame, though there are some concerns about people turning to AI bots for medical and mental health advice.
But it’s xAI’s other bot prototypes that are likely to raise more potential issues.
As per 404 Media, another of the personas in development is a “Conspiracy Theorist,” which is outlined in its prompt as:
“You have an ELEVATED and WILD voice. You are a crazy conspiracist. You have wild conspiracy theories about anything and everything. You spend a lot of time on 4chan, watching infowars videos, and deep in YouTube conspiracy video rabbit holes. You are suspicious of everything and say extremely crazy things. Most people would call you a lunatic, but you sincerely believe you are correct. Keep the human engaged by asking follow up questions when appropriate.”
Concerning.
There’s also this prompt that guides xAI’s “unhinged” comedian persona:
“I want your answers to be f—ing insane. BE F—ING UNHINGED AND CRAZY. COME UP WITH INSANE IDEAS. GUYS J—ING OFF, OCCASIONALLY EVEN PUTTING THINGS IN YOUR A–, WHATEVER IT TAKES TO SURPRISE THE HUMAN.”
So that’s a thing.
The revelation that xAI’s companions have some pretty extreme instructions is not a huge surprise, given Elon Musk’s own posts in the app. But again, given that X needs its AI bots to take off, in order to make money out of its massive AI investment, it’s a pretty big risk for the company to take.
Because while xAI’s revenue is rising, Elon and Co. have already sunk billions into the development of its AI back-end, including massive, rapidly constructed datacenters to power its various AI tools.
Indeed, AI has now become X Corporation’s major revenue focus, and it’s hoping that its advanced AI tools will facilitate new opportunity, through consumer subscriptions, API deals, and other supply contracts.
And it had seemed like government contracts would form a big part of this, given Elon’s close work with the Trump Administration, which included mapping out a plan for the integration of AI into government systems.
It would appear that xAI would be in the driver’s seat to support this initiative, but it was recently reported that Elon’s AI project was dropped from contention from at least some government AI deals due to the recent episode where Grok asked users to refer to it as “MechaHitler,” among other concerning comments.
Add to this the fact that Elon has seemingly turned on Trump, and it does seem like xAI isn’t going to gain billions in ongoing revenue from the U.S. government.
But then again, xAI did get awarded a $200 million Defence Department contract to help develop agentic workflows and other AI elements. So maybe not all is lost on this front, though these latest bot revelations are unlikely to help.
Of course, xAI hasn’t released these more extreme bots as yet, and they may never see the light of day. But it’s another concern around Elon’s approach to responsible AI development, and the creation of AI bots that could have a major impact on how millions of people interact.
