It's horrifying how many people in the UK support blocking Twitter because the AI tool is being misused. And based on an article I saw about one of the Japanese princesses, it was basically "@grok undress her" rather than elaborate prompts people seem to believe?
Given that the UK is trying to block 4chan via the scenic route, this, in my opinion, just gives the government incentive to block any website in the future and throw it back in the faces of the population.
"We're blocking $WEBSITE for $THINK_OF_THE_CHILDREN. You don't get to complain because you supported blocking Twitter, so..."
That said, I honestly believe their hatred of Elon Musk is fuelling that, rather than Grok - which very obviously needs to be fixed - itself.
https://archive.is/qssiq
This is a very interesting AI problem. Grok was trained on X content which contains a lot of porn. Other image generation models aren't trained on porn so they don't know how to produce it. It appears to be very difficult to stop Grok from making porn since it has been trained on it. There's always a prompt that can produce it. Is the ony workable solution to train a new version Grok 5 without porn?