Still working. However, I just prompted
"create an image of an orange cat robbing a bank and it replied, “there’s no response available for this search. Try asking something else.”
It worked for every other prompt. Is the “robbing” part that flagged it?
Worked for me.
Weird, I tried it twice. Oh well, here is an image it did create:
The camera must be pointing in the wrong direction (to see the robbery).
ChatGPT is a crazy good tool for creating memes. This took me five minutes to alter the original.
Trump’s casino should be run down, unlit, and have a “Gone bust” sign outside.
Like this?
ETA:
Very nice.
This actually looks better than his current UFC monstrosity.
PAARDISE.
urf.
Yikes. I didn’t notice that.
It’s on-brand, though.
It’s paar for the course for him.
ImageGPT 2.0 is on the top rankings of most of the “AI arena” testing websites. Those ones give the users a prompt and two images results from that prompt from two randomly selected models and ask “which of these follows the prompt better” or “which of these two images are better” (which are two separate questions and depend on what the particular site is trying to assess). So it’s probably the top ranking model by consensus right now, though these tests are biased towards explicit prompt following. You might think that’s the obvious thing to rank by, but I wouldn’t be so sure. Midjourney arguably produces the most beautiful images of any image generator, and sometimes it ignores parts of your prompts to do so. So they’re really just different tools rather than purely better or worse.
Midjourney, incidentally, is not included in these competitions because they don’t have an API and therefore they cannot be automated to generate thousands of images from prompts for these tests. Someone would have to do it manually. I would be really eager to see how Midjourney does, because it’s its really its own unique thing and very different from how something like ImageGPT or gemini / nano banana work. The latter are definitely better at following a specific prompt, but I find that midjourney’s images are more beautiful and the midjourney interface is a much more interesting journey of creative exploration. It’s almost two entirely different tools and workflows even though the end result is a series of images.
I generate about 90% of my images on midjourney even if the autoregressive nano banana and imagegpt 2.0 are “better.” I have a separate thread about midjourney specifically if you’re interested.
I prefer the first try.
If you’re gonna spell “paradise” correctly, at least imitate the labels on his griftphones and leave the “t” off trump.
I tend to get better results lately from Nano Banana.
I like testing models with recently-discovered obscure (to me) terms. Edentistoma octosulcatum is a species of centipede that specializes in eating other centipedes. A schultüte is a gift given to German schoolchildren on their first day of school:
A photo of an Edentistoma octosulcatum holding a Schultüte.
A couple of weeks ago was my first round of tests with the centipede.
Create a whimsical photo of Edentistoma octosulcatum driving a Corvette convertible.
I tested 7 models and I won’t include the other 6 (which had no idea what to draw) but here’s Nano Banana through Gemini:
(ETA I didn’t notice the trio of bugs in the passenger seat until just now!)
I’m sure this has been addressed, but if there’s a Trump video and he’s speaking articulately, guess what? AI!
What are you using to generate the images in all those generators? I assume some sort of router style aggregator but which one, and does it place those captions itself? Because that’s really cool
No, I made them all one by one and and combined and labeled them by hand.