Most of my recent generative GI experiments have centered on Google Gemini…which has its limitations:
“Google Gemini imposes severe restrictions against creating pictures of famous figures. You can’t create a picture of President Taylor Swift, for example.”
Why does Google impose such limits? Because it is very sensitive to misleading the public, fearful that the average person would see such a picture and mistakenly assume that Taylor Swift IS the President. In our litigious society, perhaps this is valid.
But we know that other generative AI services don’t have such restrictions.
“One common accusation about Grok is that it lacks the guardrails that other AI services have.”
During a few spare moments this morning, I signed up for a Bredemarket Grok account. I have a personal X (Twitter) account, but haven’t used it in a long time, so this was a fresh sign up.,
And you know the first thing that I tried to do.
Grok.
Grok created it with no problem. Actually, there is a problem, because Grok apparently is not a large multimodal model and cannot precisely generate text in its image generator. But hey, no one will notice “TWIRSHIITE BOUSE,” will they?
But wait, there’s more! After I generated the image, I saw a button to generate a video. I thought that this required the paid service, but apparently the free service allows limited video generation.
I may be conducting some video experiments some time soon. But will I maintain my ethics…and my sanity?
Bredemarket 