I was just messing around with an art generator at Night Cafe. Seems pretty simple.
Has anyone done any Traveller art with the AI tools out there?
Has anyone done any Traveller art with the AI tools out there?
General Has anyone done any Traveller art with the AI tools out there?
- Thread starter Spinward Scout
- Start date
maybe even a do-si-do! Rumor does have it that Firefly was Whedon's college Traveller game.
"Is anyone using AI for Traveller?"
Umm ... sort of. I find the AI great for generating inspiration that I then need to rework the Art myself. Here is an example. I gave the AI this start image ...
Then I asked the AI to create a Cyborg Pet from that image and I got THIS:
Now that is NOT what I was asking for and it needs some work, but that is an idea that I can work with. It inspired me.
Umm ... sort of. I find the AI great for generating inspiration that I then need to rework the Art myself. Here is an example. I gave the AI this start image ...
Then I asked the AI to create a Cyborg Pet from that image and I got THIS:
Now that is NOT what I was asking for and it needs some work, but that is an idea that I can work with. It inspired me.
FreeTrav
SOC-14 1K
Many AIs seem to weight phrases in the prompt based on their location within the prompt - that is, the earlier mentioned, the more important. As such, place the most important descriptives up near the beginning of the prompt. For example, "a green-skinned humanoid with pointed ears wearing a military uniform" is more likely to get you a Vulcan, perhaps not in uniform; "a military officer with pointed ears and green skin" is more likely to get you something that could pass as an Imperial officer, who may or may not be able to pass as Darrian or as Vulcan.
(It will also depend on which generator you're using; Wombo doesn't seem to do people as well as NightCafe, for example.)
(It will also depend on which generator you're using; Wombo doesn't seem to do people as well as NightCafe, for example.)
Well, I am tinkering with Night Cafe as atpollard and all of my prompts are visible.
My best advice, for a game (not necessarily publication) is to find a good "Start Image" that has the basic form of what you want and then evolve it using the AI. I started with a basilisk illustration because I wanted "8 legs" and evolved it to be more "Bear/Lion/Wolf" in appearance. The final result is posted among my work on Night Cafe. [It was for my PbP game.]
"Start Image"
First AI Evolution:
Final AI Evolution:
Last edited:
Timerover51
SOC-14 5K
That looks really good. You are further along than I am with using the various tools. For those interested, I post as Wildlife51, and all of my prompts and setting are visible.
I have been working with a fewI use it for player and NPC character portraits - since I'm bad at art (both digital and manual) and stealing from the internet gives such a wide variety of art styles.
Using an AI program with the same basic art style selected allows me to have all the characters look like they come from the same universe (art style wise).
I've attached the player characters for my most recent Traveller campaign (the first I've run offline in Korea).
Using an AI program with the same basic art style selected allows me to have all the characters look like they come from the same universe (art style wise).
I've attached the player characters for my most recent Traveller campaign (the first I've run offline in Korea).
Attachments
Wait till you see the scorpion gunship.
The Vargr: "Portrait of a friendly wolf-like humanoid alien with completely black fur in a science fiction setting" was the original prompt. But Playground modifies the prompts based on a chosen style. For example, it changed my original prompt for the Zhodani to:
"Acrylic painting of a poised, athletic Persian woman aged 40 with a turban and futuristic attire, standing 185 cm tall in a sci-fi environment, inspired by the styles of Makoto Shinkai, Jamie Wyeth, James Gilleard, Edward Hopper, Greg Rutkowski, and Studio Ghibli, with a blend of Genshin Impact aesthetics, utilizing both palette knife and brush strokes"
These expanded prompts also show a big problem IMO: AI does not in fact create much. It copies and remixes - including works of artists (like Greg Rutkowski in this case) who are outspokingly anti-AI art.
I would not be comfortable using this in a published work (whether for free or especially not for money) - I use AI art mainly for quick(ish) NPC portrait generation and mood-setting pieces such as the one in the following post.
"Acrylic painting of a poised, athletic Persian woman aged 40 with a turban and futuristic attire, standing 185 cm tall in a sci-fi environment, inspired by the styles of Makoto Shinkai, Jamie Wyeth, James Gilleard, Edward Hopper, Greg Rutkowski, and Studio Ghibli, with a blend of Genshin Impact aesthetics, utilizing both palette knife and brush strokes"
These expanded prompts also show a big problem IMO: AI does not in fact create much. It copies and remixes - including works of artists (like Greg Rutkowski in this case) who are outspokingly anti-AI art.
I would not be comfortable using this in a published work (whether for free or especially not for money) - I use AI art mainly for quick(ish) NPC portrait generation and mood-setting pieces such as the one in the following post.
Last edited:
"Setting the mood" illustration for a campaign I'm currently running.
Prompt:
"Panorama of a small sensor station built into large rock needles about 50 meters in height in a swamp surrounded by a dense jungle of blue and green plants on an extraterrestrial planet, covered by mist."
As you can see, much of the prompt was ignored, but it came out well nonetheless, so I used it.
I have always found the turban depictions rather cliche not that your attempt is wrong, But I always wanted to have my Zhodani in the Vietnamese style turbans. Still Haven’t got it right
Attachments
Last edited:
