Grok Does Images Now, and People Are Angry

But then, when aren't they?

If somebody paints a mean picture of you, who do you get mad at? The person, or the paintbrush?

I’m pondering such imponderables this morning because Twitter¹ just updated its Grok² AI to generate images. Except unlike most other image generators these days, Grok doesn’t seem to have any guardrails.

Yet.

For example, a lot of image generators will balk at a text prompt like this:

Bill Cosby, wearing one of his famous sweaters, laughing while picking up a prescription at a pharmacy. The pharmacist is an attractive woman.

Give most image generators a text prompt like that, and they’ll hock up an error message for whatever reason. Because it’s “morally wrong,” or the company “doesn’t want to get sued,” or whatever. They choose to censor such an image, because they don’t want to get blamed for it.

They don’t want you having any fun.

But Grok really doesn’t give a $#!+. Grok does whatever you want: