So, if wordseye.com/ can be made to plug into Inform 7, it will suddenly become a graphical adventure creator! Scribblenauts was only the beginning…
Umm this is fascinating.
Í must have a horrible mind, but all I can really think about is the limitations of this system.
But busterwrites’ adjective is the best - it’s exactly what it is.
Hey, this provides everything an author of AIF will ever need.
The giant T-Rex is shooting lasers out of its eyes.
A robot android ninja assassin wearing sunglasses and wielding a laser katana and machine gun is riding the T-Rex.
Giant squids are attacking from the sky and are getting shot by the T-Rex and the android ninja assasin.
Around all this, pink floppy unicorns are dancing on rainbows.
That’s how I read that sentence.
Now I feel kinda disappointed.
Could someone with an account put this in and tell us what comes out? 
(I suspect it’ll complain like I7 does when people actually think it’ll magically understand what it’s saying. But why not give it a whirl?)
If only we could use this to auto-generate 3D maps…
You imagine an amazing true to life map of Anchorhead, the Colossal Cave of Adventure, the endless domains of Endless Nameless…
But it would just be a load of boxes with generic labels too tiny to read - the same as regular 2d maps.
Not necessarily, if this could parse out the scenery from room descriptions.
A lot of IF scenery is invisible unless you refer to it in a command. I wonder if Quest game files tag scenery?
Thank you for your interest in Wordseye! You can click on this link and get immediate access:
wordseye.com/registration?b … iction-Nov
We are excited to hear what you think and look forward to seeing your posts in the Gallery!
Gary Zamchick
type-a-picture@wordseye.com
www.W\wordsEye.com
Wordseye is more a generative art application than an AI artist which can make words into a perfect picture in one try. The default output needs a human to zoom/pan to target a subject and to apply effects. Objects aren’t sized to be compatible (the library is a hodgepodge) so you will need to specify attributes to fix their sizes.
I tried to make specific images but the free library of 3d-models and textures isn’t big (the model is that you can pay for access to more) - if you’re serious then you bring your own models and textures - but then if you’re serious why not use professional art and 3d tools which offer real-time feedback?
This web app is free and If you have lots of time then it’s fun to play around and see what cool images result. I don’t see any scenes in the gallery with 100 objects, I guess it’s aimed at sharing a quick idea on social media.