Ai models like Grok have zero human intelligence, which is especially obvious when "chatting" about real world events. These models know about information from one year ago.
Each training cycle of an Ai takes in huge volumes of misinformation and lies, requiring a lot of human intervention to elevate it from a moron-bot repeating lies and rumors.
A (24 hour/7 day a week) live training model would need to gain actual human intelligence, which none have, or require a huge number of human assistants.
Grok says that videos of Charlie Kirk being shot are memes and that he's actually fine