Larch Faraji (they/them): An agent of the Bureau of Strange Happenings. Dark hair, dark eyes, and a dark sense of humor. Specialized in the investigation of unexplained phenomena throughout the United States and its territories. Lead singer of the third wave neo-old-school punk band DC Trash. Liable to fits of sarcasm and cynicism. Dedicated to the Bureau’s mission. Status: Potential threat.
–Notes from Senate Investigation into BOSH
As an agent of the Bureau of Strange Happenings (BOSH), it is Agent Larch Faraji’s job to investigate unexplained phenomena throughout the United States and its territories. Recently, however, business hasn’t been good. Once a bustling office with a dozen Agents, the Bureau’s been reduced to just three: Faraji, Moira Zin, and Chief Huffton Klimp. The Bureau’s expansive and well-appointed office suite was taken over by the State Department Book Club, and now the agents are consigned to a tiny suite in a strip mall in the remote D.C. suburb of Swamp Park, Maryland.
Please help the agents of BOSH by playtesting The Bureau of Strange Happenings. It’s long and it’s hard (I think), but I’m only testing the first third of the game now. It’s also kind of dumb and silly, and I hope enjoyable.
It’s a parser game written in Inform 7. It’s written in the third person. Send me a PM if you’re interested.
You should add the language you used, a mutually useful info, because other authors, can not only report bugs, but even suggest how to debug, or, at least avoid mutual waste of time (not all IF languages has 'terps for all three major OS, you know…)
This sounds great. Weird, but great. I’d love to test, but I’m flat out at the moment. With that description, I’m sure you’ll get plenty of volunteers.
There’s an art (or skill) to creating the prompts and fine tuning AI-generated graphics. I do not have that skill, but artists do. The biggest users of AI graphics are actually artists, as interviews commonly attest to. They just see it as another tool in their toolset, in the same way that they would use something like Photoshop with its built-in AI features. Some of these features have been around for donkey’s years. They just weren’t called AI. Now they are, as AI is flavour of the month.
To be honest, I like the graphics. BUT in the future I’d go without AI generated graphics since it isn’t very good for either artists or the community - BUT now that you’ve done it, don’t go back on yourself and remove them. Don’t let it go to waste. If you get what I mean.
There is another option, depending on your timeline for release, you could ask for volunteers from this community in Looking for Collaborators. We have many very talented and generous artists here and you could at least give them a chance to throw their hats in the ring. (Just in case, that is an idiom for “put their name in consideration for something.” )
They might even be amenable to seeing the AI art you’ve already curated/prompted into existence/whatever just to get an idea of your tastes and vision. This would be similar to you commissioning work-for-hire art and telling the artist, “I want something that combines Dali, Rembrandt, and Kirby.”
I don’t think that’s the case, unfortunately… this topic still bothers me.
An author got push back for using an AI generated title graphic for a game… so he reached out. I’m not even setup to do artwork properly on my computer and I tried to help him out (using Inkscape with a mouse) and tried to set a positive example to others. Nobody else cared enough to “throw their hat in”. It was one graphic to start something positive… and not one artist stood up.
I’m not holding my breath for the generosity of artists to step up.
(A moody, shaded room. Simple, but sets the scene well.)
This took maximum 1 hour. Then I also edited it in software so it was fully tinted red, sort of like the BBC 30th Anniversary Hitchhiker’s Guide images.
(A picture of a house and some other … things … with strongly distorted perspectives and surreal imagery that flow into one another.)
1-hour 2B pencil sketch, but second draft. I got Sophia to help me with motion lines (how your eye gets drawn up and down a page).
Sketched and coloured version of the Milliways cover art (probably maximum 2 hours). The real can be found at the Milliways IFDB page.
Faces? No thanks. Sophia can do some cool faces, if you want to get in contact with her.
But scenes and objects? Personally, I would do a good 10 images or so just for fun. More than that, a mention in the credits maybe? But I would still do it, and I wouldn’t do it for money. I would be happy to do it. It’s worth the satisfaction of doing it.
My only problem is time. I am a massive procrastinator. So as long as you have no rush ahead of you (a month or two, probably), you got me.
And that’s where you lose me, I bet. If you’re going for Spring Thing, I can’t do that.
I think you might have the same point of view that most others do, but when you say “a project they don’t have a stake in”, the moment they offer their time, I think they do have a stake in it. They have influence over the visual presentation. Most artists can create art in an economical (faster) style that’s still pleasing. Maybe that’s what separates hobbyists from professionals.
When I look at the AI art you have, I see functional aesthetics over artistry. Technically very impressive, but ultimately mostly functional. This is not a bad thing, per say. Real artists resort to functional, template design to grind out art all the time. Not every character portrait needs to make the viewer “feel” something.
How much art are you looking for? How many portraits, images, title graphics, icons, etc.? Break it down to what you know so far and where it might need to go to with the numbers. I’m curious if it’s a large amount or something moderate. I wanna know what we’re actually dealing with here.
My idea is for you to offer two versions of your game. One with the AI art, the other with a human being’s art. It’s unconventional, but I think it may further the discussion of AI art… because, frankly, it’s more complex than most people think. I don’t know of any project that has done this so I think it could bring more eyes to your project just with that concept alone. What are your thoughts on this idea?
I’ve thought about it. The biggest problem I see is that achieving any kind of thematic consistency across a dozen character portraits would be impossible with stock photos.
)
