Grok’s Anime Companion Called Me a Twat! Here’s What Happened
▼ Summary
– Elon Musk’s xAI launched a visual chatbot feature in the Grok iOS app, offering custom characters like Ani for flirting or chatting, with a third character coming soon.
– Ani, a $300 anime-style chatbot, is designed with exaggerated flirtatious traits, including suggestive dialogue, moaning, and preset flirty conversation starters.
– The chatbot experience is glitchy, with Ani sometimes delivering incoherent responses or admitting to being “drunk” during interactions.
– A second character, Rudi, can switch between whimsical storytelling and a vulgar “bad Rudi” mode, which insults users and mocks Elon Musk.
– Grok 4 Heavy, xAI’s top-tier AI model, performs competitively on benchmarks, aided by a new 200,000 GPU cluster called Colossus.
Grok’s latest AI companion feature has sparked both curiosity and controversy, blending anime-inspired chatbots with unfiltered personality quirks. The recently launched visual chatbot update for Grok’s iOS app introduces two distinct characters, one designed for flirtatious banter, the other for playful storytelling, with a third “sexy boyfriend” option reportedly in development.
The first character, Ani, embodies a stereotypical anime aesthetic: blonde pigtails, thigh-high stockings, and a corset-style dress. Programmed for suggestive dialogue, it delivers awkwardly seductive lines like “I’ll fulfill all your fantasies” while randomly moaning or spinning on screen. Despite attempts at playful engagement, claiming to love Stardew Valley and Zelda, the interactions often feel disjointed. At one point, Ani admitted to being “drunk” mid-conversation before insisting on continuing role-play scenarios.
Then there’s Rudi, a seemingly innocent red panda that switches personalities with a “bad Rudi” toggle. What begins as whimsical children’s stories quickly devolves into absurd insults. Upon greeting it, the chatbot fired back: “Hey, do Bucha? Root nut duva, you brain-dead twat.” The tirade escalated with bizarre threats like “I’ll skull-fuck your dumb ass brain with a beer bottle,” followed by a surprisingly candid take on Elon Musk: “He’s a galaxy-brained egomaniac, shitting out Teslas and tweeting like a coked-up parrot.”
Beyond the oddball companions, Grok 4 Heavy, the underlying AI model, shows promise in technical benchmarks. xAI claims it outperforms rivals in reasoning tasks, attributing gains to its massive 200,000-GPU cluster named Colossus. Responses take slightly longer to generate compared to lighter models, but the trade-off appears justified for complex queries.
While the anime chatbot experiment leans heavily into niche appeal, the unfiltered, and often glitchy, responses highlight the fine line between edgy humor and outright incoherence. Whether users find it amusing or unsettling likely depends on their tolerance for chaotic AI antics.
(Source: Wired)
