site banner

Small-Scale Question Sunday for October 8, 2023

Do you have a dumb question that you're kind of embarrassed to ask in the main thread? Is there something you're just not sure about?

This is your opportunity to ask questions. No question too simple or too silly.

Culture war topics are accepted, and proposals for a better intro post are appreciated.

2
Jump in the discussion.

No email address required.

Does anyone have access to ChatGPT with images? Is there some opt-in setting I have to find and enable, or are they still doing a staged roll-out?

I have Dalle3 accecss. It's... alright.

Humans look extremely uncanny. All pictures look like 3-d renders.

Worst of all, They reword your prompt in 4 different styles.. why? It's even funnier when you want pictures of people. It will automatically give you 4 different races (evident in the prompt lol).

I'm really not liking how much OpenAI tries to baby its users.

We using the same DALL-E 3?

Because for me, it's amazing, almost solving the problem of AI image generation barring truly baroque prompts.

It can be as aesthetic as Midjourney, while having far more intelligence, especially for semantics and interactions between objects, than either that or SD.

Previously, asking for artwork of a woman with cyborg legs holding her child while they curiously look up at a painting of a superhero would have produced less than ideal results, like a female cyborg child looking at her mom or something. It usually works these days.

And humans look human, in every art style I cared to try, which is an improvement from the previous DALL-E 2 Experimental, itself far better than the first DALL-E 2.

Here, pictures are a thousand words or something along those lines, depending on the compression algorithm I bet:

And I'm cherry picking the best out of 2 or 3 acceptable images in a batch of 4, not one from 4 batches like I had to previously

I guess it's only natural that AI would generate a picture of Internet Explorer Man embracing the older generation.

Ouch.

If you'd like to know what I was trying to achieve, that guy is a knockoff Superman who is simultaneously embracing his mother as she actually is, young, healthy and beautiful after anti-aging therapy, contrasted with how he sees her in his mind's eye, old and withered, as would any woman of her age be today.

Largely a psychological hangup from the fact that his superpowers leave him unable to avail of some of the options available to baseline humans, such as said therapies, cybernetics and so on. He has a hard time accepting that despite the entire world looking up to him, he's going to age and die in a few decades, while she might live on for centuries, millennia, or till the Heat Death of the universe. Well, maybe not, if the hostile K2 aliens next door have something to say about that. But in theory at least.

Dude, that granny's arm in the first pic is SWOLE. Damn granny.

HGH is a helluva drug! Maybe that's where the superhero, Consul, got his.

The proportions are still really weird if you look closely. And I agree that it had the kind of waxy, Digital Art^TM look that the other people mentioned, I want one of these models that’s only trained on scanned physical art, just to see what changes.

Waxy? I don't really see it, but I use a combination of artist names and keywords that have served me well since before SD launched for the public.

You can get more naturalistic results if you want, I just wanted this look.

Sorry I mean the ability to upload images, e.g. upload a photo of a menu with the prompt "pick something vegan".

Did you have to do anything, or did a Dalle3 button one day appear for you on the interface?

Staged roll out possibly dependent on account history and region.

I follow Mikhail Parakhin, so I know a full rollout was done for the version in Bing Image Creator a few days back.

Ah, I’m talking about ChatGPT Plus.