Interactive Fiction: Story-Wrestling with AI Dungeon

Someone on Mastodon shared out that a new interactive fiction game app – AI Dungeon — was free for a limited time (still is, as far as I know, for now), so I took the plunge and tried it out. It’s quite fascinating, as the AI software brain spins an elaborate story from your text-based responses, questions, actions. There were some basic themes built into the game but I decided to test out a more open-ended option and set my story of a musician/spy into motion.

Here was the initial story set-up I set into motion:

You are a traveling musician who has been sent to spy on a neighboring kingdom. The road you travel brings you through many small villages. You meet many other musicians, and maybe more than a few spies. What do you do?

As I understand it, after a bit of research and computation, the AI brain sends story narrative back based on its interpretation of my actions within the game. Hints in the game suggest that more complex the user language/words are, the more the AI will learn and adapt to the story.

Interestingly, there were definitely times when I could feel the AI tugging my story into its known corners — I ended up in a cave a few times — as opposed to truly letting my activities guide the story forward on its own path. The AI writing itself was remarkable coherent, for the most part, although sometimes, when it either provided dialogue or used mine, things got a bit convoluted in the context of who was speaking.

The developer site explains (and deeper technical explanation here):

AIDungeon2 is a first of its kind AI generated text adventure. Using a 1.5B parameter machine learning model called GPT-2 AIDungeon2 generates the story and results of your actions as you play in this virtual world.
Unlike virtually every other game in existence, you are not limited by the imagination of the developer in what you can do. Any thing you can express in language can be your action and the AI dungeon master will decide how the world responds to your actions.

It took me some time to find a rhythm of my own, to stay true to my sense of character (actually, I’m still doing that as I keep playing) — I’d have them (me?) pull out a guitar or saxophone or harmonica, now and then, and use music to discover mystery — and not let the AI be the one in charge of the story. It was a bit of story-wrestling, in a fun way.

At this moment (as I write this blog post), this is what happened after I played my saxophone:

You play the saxophone, which causes the water around you to ripple slightly. Suddenly, a bright light shines through the hole in the rocks. You look out into darkness and see a tall figure emerge from the cave entrance. It’s face is covered with black hair and its skin is pale white.

After a few mornings of “playing the story,” I still had not gotten too far into where my character was actually going and trying to do (other than spy), and why they were going there, although I had met my fair share of interesting characters (again, some seemed to have been yanked from some other game world into my own) and entered some intriguing rooms. I broke mirrors, used keys, sent messages via guitar string, ran from one person and found another person, took a horse for a ride, and more — all with text-based storytelling, guided by AI database.

I’d love to see where AI Dungeon goes. It’s still being developed and the brothers behind the company hope to fund their project through Patreon. I can’t afford the $5 month, their lowest tier (I wish there were something even lower, but I feel like a cheapskate even suggesting that), but if there’s a way to keep playing and supporting a version, I’ll do it.

You can read my story from a link generated by the app (you don’t seem to need the app to read the story, which is helpful).

Peace (playing it),
Kevin

The Only Reliable Tech is Pen and Paper

This is becoming a regular story of mine, and probably yours. An app that I really liked using for writing and making digitally, and on which I relied upon regularly, seems to have gone dead on me. Its name is Legend, and it was an animated text app that Terry Elliott turned me on to long ago.

And I loved it on my iPad, for its simplicity and its design and the way you could easily find Creative Common images via its Flickr connection and then layer short text on top of the image. I used Legend for poetry and for quotes, and for merging words with motion and image. It was my go-to app for many things.

And now Legend is gone from the App Store. Vanished without a trace.

I was having some troubles with Legend on my iPad the other day, and I deleted the app in hopes of re-installing and re-booting it, and soon discovered that the app itself was nowhere to be found in the iTunes App Store. It’s not even a mention anymore in my “bought” apps file bin in iTunes. It’s like it never even existed, and its loss saddens me.

But, of course, I should know better.

In this world of digital writing and composition (and art, and whatever else we want to call it) the only technology that really stands the test of time with any consistency is a piece of paper and a pencil or pen. All else is mostly temporary, so be sure to back your stuff up and keep an eye on the horizon for alternatives.

In my mourning for Legend, I have been trying out a few different animated text apps. I grabbed a quote from Dave Cormier’s recent piece on Rhizomatic Learning.

This is HypeText app:

Experimenting with Animated Text Apps (w/Dave Cormier quote)

and this is TypiVideo app:

Experimenting with Animated Text Apps (w/Dave Cormier quote)

And this is TypoTastic app:

Experimenting with Animated Text Apps (w/Dave Cormier quote)

None do what I want it to do. None feel quite right. Some have limits on loop time (either going too fast or too slow). Some don’t give you much access to images beyond your own files (which has value but requires deeper planning than I am usually doing for this kind of work.)

I’ll keep exploring. I am checking out Legend on my Android phone … hmmm … seems like it now has an entirely new name now (Animated Text), and has advertisements within it … and no longer has access to Creative Commons images. Dang.

This exploration is another reminder to myself, and maybe to you, that nothing lasts forever in this shifting environment of operating system updates, app development, and that our own means and venues of digital writing is always in flux and motion.

Peace (animated with image),
Kevin

PS — I’d also like to say that I could probably do what I want with animated text via Keynote or Powerpoint (and I have) but I appreciated the ease of making animated texts with Legend and other apps. Maybe another post for another day is about what we give up as writers — creative control, freedom to make change, a vision from start to finish — when we allow our tools to guide our writing process.

Aw Snap — Introducing Digital Annotation with #BookSnaps

BookSnap Mentor Example

I ran across a reference to an idea called BookSnaps that seemed intriguing so I followed the thread to Tara Martin’s blog, where she shared out information about how to use digital tools, particularly Snapchat, for annotation and layering of media.

Watch Tara’s short talk/presentation about the idea:

I was intrigued because I am interested in finding more ways to engage my sixth graders with annotation and digital tools, for many of the reasons that Tara gives: the ways annotation focuses attention, how it helps us remember, how to it makes visible the learning of a text.

While Tara shares about Snapchat as the platform, I was more interested about using something within our students’ Google accounts, to make it easier to teach and easier to save. We are in our Independent Reading unit right now, so this is a perfect way to share the first pages of books they have chosen, I am thinking.

My sample — for Philip Pullman’s The Book of Dust, see above — was done in Google Drawing and it all went quite well, using call-out text boxes for the writing and some images searches for the “stickers.” There’s not a lot of space, so finding focus will be key, as will setting parameters for how many overlays can be on a page. I can see my kids getting carried away with images.

Tara does have a video about using Google Drawing that helped me think this through:

(Note: Google has now changed the way one can take image snapshots within its system, so the direct method that Tara mentions in her video may no longer work. I used PhotoBooth for my sample, but Tara kindly mentioned a free extension by Alice Keeler for Chrome that takes pictures and puts them into a Google Drive folder, which can then be moved into Google Drawing. I tested it out and it seemed to work quite well.)

I envision this BookSnap idea as one of the first steps of our work with digital annotation, and the connection to Snap Chat (even though we won’t be using it) with layered text and layered image, and sharing, should grab my students’ attention. And sharing out books, and reading about what others are reading, is always a powerful sharing experience, made more fun with layers of annotation.

I’ll let you know how it goes …

If you are thinking that the use of Snapchat App is of interest, this video by another teacher (not Tara) gives a good walk-through of each step along the way:

Peace (layer it and annotate it),
Kevin

Make Panoramic Virtual Reality in Cardboard

Cardboard VR

Thanks to Richard Byrne, over at Free Tech for Teachers, I began exploring Google Cardboard again, but this time, I used the Google Cardboard Camera app to make my own virtual reality. The free app allows you to create a 360 degree panorama, and you can even add audio narration, and then share it out with people. (I did it with my Droid phone but I believe you can do the same with any iOS device.)

I did a trial run in my backyard and then used it when I did a hike up to a fire tower known as Goat’s Peak, which gives you a nice view of the Pioneer Valley of Western Massachusetts from top Mount Tom.

If you have Google Cardboard (or some VR goggles .. I bought my set from Amazon for about 10 bucks) and the Google Cardboard viewer app, the links should open up for you in that app. You can see what I saw and hear me explaining it a bit.

See my backyard

See Pioneer Valley from Goat’s Peak

Here is Richard’s very helpful video tutorial on making and sharing Google Cardboard VR.

It would be pretty cool to have students use the Google Cardboard Camera app to make their own virtual reality shows of special places.

Peace (it’s reality),
Kevin

A SmallPoem for Small Poems

I’m tinkering around with a visual typography app that Terry and Wendy shared out called TypiVideo. I like the effects of the moving text but I am having trouble with finding the ways that I can set animation and text (I know the controls are there, and I saw a tutorial that indicated where and how I can do more, but I can’t seem to have get to them to work on mine. Might need to reload the app.)

Anyway, the poem above is for another poetry venture elsewhere.

Peace (animated),
Kevin

If I Were the App Designer for the Museum …

Design App for Armory

At our summer camp project at the Springfield Armory, as our middle school students were working on a variety of projects, so were us teachers. I had already shared out my Rosie the Riveter 2.0 project. One of the Armory rangers had mentioned that another ranger is in the midst of planning out an app for the museum. I decided I would try to imagine what might be in a Virtual Reality  app for the Springfield Armory, which is a National Historic Site for its role in our country’s history.

The ranger asked me to leave the drawing behind, in case it offers up ideas for thinking about the museum in the future as they begin to mull over ways to make the museum more interactive. So, who knows? My ideas might someday become reality. Or not. My favorite, and the kids’ favorite too, is the virtual roller coaster set within the gears of the innovative machinery of the Armory.

It was fun to think in terms of Virtual Reality.

Peace (make it real),
Kevin

Slice of Life (Day Five): On the Possibilities of Collaboration

(This is for the Slice of Life challenge, hosted by Two Writing Teachers. We write all through March, every day, about the small moments in the larger perspective … or is that the larger perspective in the smaller moments? You write, too.)

One of my hesitations in jumping into Slice of Life is my participation in something known as Networked Narratives, a ‘course’ being run at Kean University by Mia Zamora and Alan Levine (remotely) which has an open online invitational component, which I am part of. So, this Slice of Life sort of converges with Networked Narratives. That’s a good thing.

My good friend, Wendy, from Australia, has been tinkering with the app Acapella as a way to foster more narrative collaborations with the NetNarr folks, mostly those of us out here in the wide open spaces. The students in the actual course seem a bit more restrained and follow the course’s activity guidelines pretty closely. Out here, we just do what we wanna do. We’re not getting graded, of course.

Anyhoo … Wendy and I have been trying to navigate the potential of the Acapella app, which has strange quirks around collaboration yet has some potential that we find intriguing enough to stay with it. We’ve messaging back and forth, working through the kinks and frustrations.

This is one of our impromptu collaborations.

Next up is an invite to a few more friends (Sandy and Terry) and plan out something a little more creative and focused.

Peace (in collaboration),
Kevin

PS — this is one acapella mix I made myself long ago, when I first tried out the app.

App Review: StickNodes

A few weeks ago, for the #CLMOOC DigiWriMo Pop Up Make Cycle, the focus was on animation. There are all sorts of apps that allow you to animate now, and StickNodes is one of my favorites (I paid the $1.99 for the Pro version). It’s an update on an old freeware that I used to use with students called Pivot Animator. When we shifted to Macs, I had to move away from Pivot (it is a PC-only freeware) and tried Stykz for a bit.

StickNodes Pro is pretty easy to use, and has a lot of powerful features for animating stick figures. It’s also pretty darn fun to use. You can create and then export your animation as video or gif files, which can be hosted elsewhere.

Here is one of my early experiments: Stickman Walking. (I had uploaded it into Vine, which you can no longer do)

 

No surprise that there are tutorial videos on YouTube for using the app. Here is the first in a series done by this person.

Give it a try. Or try some other app, and let us know. We’re animating this week!

Peace (in the frame),
Kevin

The Appeal of Musical.ly (and the Trap of the Likes)

musicaly

My 11 year old son has really enjoyed making videos with the Musical.ly app this summer. If you don’t know what it that is, Musical.ly is a lip-syncing app, which provides short bursts of song that the user creates an equally short video for. Most people lip-sync the song and share within the community. Tons of kids are using it.

I was going through the videos my son had been making (nearly 100 now), and I was laughing at his sense of humor. His friend was over this weekend, and they made a very interesting one, that was shot entirely in reverse, as his friend catches some inflatable baseballs and items while grooving to a song. A sort of trick-camera thing. I thought it neat, but it might just be a simple tweak of the app. I’m not sure.

(Note: I asked later the boys about it. It was done in the app. So much for hacking the app for creative video editing.)

I also get a little antsy, though, in how my son, entering his middle school years, gets caught up in the number of viewers and likes and all of the suitcase luggage of social media (thumbs up, plus, etc.) on his work (not just here in this app but also in other platforms) that doesn’t really designate anything much in reality. He sees it as “someone is watching what I am making” but I suspect it is an ego thing, too. He wants to be popular, and he sees technology as one of the ways to be “cool.”

We’re already overhearing conversations about “how many views do you have?” and “how many videos have you made?” as if it were all a cold numbers game.

We’re trying to temper that impulse for “likes without context” with discussions at home, as best we can. He’s always enjoyed making movies, and has regularly written and produced videos himself with friends. (In fact, he is finishing up the editing a video that he and this same friend shot over the weekend.)

This Musical.ly app is designed for short videos, and I hope that it doesn’t suck dry the creative fountain for his desire to make longer video productions. I hope, instead, it gives him more ideas that he can use elsewhere.

And sure, he’s having fun with it. That’s important, too.

Peace (in navigating the world),
Kevin

App Review: Gum

I am intrigued by ways in which social media and technology bridge the gaps between real space and virtual space. So, augmented reality apps are interesting (if still a bit complicated to use). Virtual reality ideas pique my interest. So when I saw someone share out about the app, Gum, I thought I would give it a shot.

Gum is an app that uses the bar codes on products (such as food) as a means to leave comments and texts related to those products. So, for example, if I open up the Gum app and connect it to the bar code of my older son’s favorite food — Ramen Noodles, chicken-flavor — I can leave some thoughts about the noodles. Anyone who uses the Gum app and scans in the bar code of Ramen Noodles (chicken) will now see the comment I left there about alternative uses for Ramen noodles.

And they can leave their own comment, too.

(Try scanning the bar code here of my Ramen Noodle package)

I left a question on the other kid-fave food in our house — Annie’s Mac and Cheese (purple box). Maybe someone will answer it in the future, using the Gum app. I’ll get a message if someone does. (I hope someone does).

(Try scanning the bar code of my Annie’s Mac and Cheese box)

I had thought, initially, that I could maybe upload an image with my text (and even crafted a comic for the noodles), but I guess it is just text connections across social media space with real objects. I’m still interested, and wonder how a network like CLMOOC might tap and hack into this kind of app for some connected project. How might we riff together with a connection to physical space with Gum?

Peace (yum … gum),
Kevin