Generative AI Presentation 3: Ethics, Policies and Reaching All Students

I did one final summer workshop presentation about the rise of Generative AI and the impact on schools. This one, like my last presentation, was with a group of school librarians, who were doing a week-long professional development around reaching all students, including those with disabilities.

My themes for this presentation centered first on the ethical considerations of using AI like ChatGPT and Bard, and then moved into how to begin thinking of school/classroom/libraries policies for students, and then ended with some ways that new AI-themed tools could help educators support students with some disabilities. I then provided a wide range of resources for them to explore and consider. (By the way, Eric Curts has done an amazing job in this area of inquiry, gathering resources on this topic.)

This is the third workshop I have given this summer (first with teachers in the Western Massachusetts Writing Project and then with another group of school librarians), and interest in all three has been very high by participants. We’re all wondering how AI is changing the educational landscape, and how we can adapt to meet the moment. Even for educators whose students are too young for ChatGPT or other tools, the interest is there to learn more.

I’ve tried to balance the concerns we all have with an open mind to the possibilities the AI might bring to the classroom. Threading that needle in these conversations can be tricky, but mostly, I have found, when educators learn more through discussion and explorations, the more open they are to considering the possibilities along with any pitfalls.

Our Western Massachusetts Writing Project is starting to plan a gathering of educators in the fall to do more of this work around AI, and I am likely facilitating part of that discussion with a professor from Westfield State University, our new home for WMWP.

Peace (and purpose),
Kevin

Generative AI Presentation 2: School Libraries and ELL Students

This summer, I have been doing a handful of workshops with educators on the topic of Generative AI. A few weeks ago, for example, I visited the Summer Institute for the Western Massachusetts Writing Project and facilitated a fascinating inquiry into the rise of AI in education.

This week, I worked with a group of school librarians, exploring what Generative AI is, some ways that school librarians might consider platforms like ChatGPT and Bard and others, and a focus on how to reach and support English Language Learners (a focus of the week-long professional development – led, in part, by my wife –  I was invited to present at).

I’ve been trying to open these sessions with impressions of educators, and in this case, I used Answer Garden (I was presenting remote) and, as in past sessions, the responses are interesting, running the gamut from alarm to interest.

School Librarians: Impressions of AI

View the basic presentation in full-screen mode.

Next week, I will be presenting to a similar group, also of school librarians, with a focus on the ethical considerations of AI, policies for schools and classrooms, and how AI could support students with learning disabilities (the focus of that week’s PD with librarians).

I am no expert in AI, by any means, but I am finding my deeper personal inquiry into these platforms is helpful for other educators, and I am enjoying the explorations of this new emerging world with others. And I am trying to work into these sessions opportunities for colleagues to play, explore, learn, and reflect — even if they don’t think they will be tapping into AI anytime soon (for a variety of reasons — privacy, access, age, etc), it’s important to have a basic grasp on what’s happening in the field of Artificial Intelligence — for our students surely have some knowledge, if not experience, themselves.

Peace (and presentations),
Kevin

Video By AI: Saxophones, Dogs, Flowers and Faces

A neighbor of mine, after a street discussion about the potential and danger of AI to the world of art, suggested I give Runway AI a try, as it turns text and images into movement videos. I know this kind of technology (word to video) is still in early stages, so I was not all that surprised by the weirdness of the results. And yes, it was odd, all right.

I first started by uploading an image of our dog with a ball toy but it was so strange. Her face got really contorted and turned into another dog altogether (not nearly as cute!).

My next experiment was with text prompts. First, I asked for a saxophone player on a jazz club stage at night. I don’t know what kind of saxophone this is or why he is playing it with his nose, but … I was amused. (see above)

Next, I asked it in text for Wind Over Field of Sunflowers, and it is rather lovely. The flowers have an oversized head, cartoon-y look to them, and they move ever so slightly in the wind in a slow video pan. This one was the best of the bunch.

Finally, I uploaded an image of a face made in sand at the beach, and got … more weirdness. I don’t even know how it got to that ending.

Peace (and play),
Kevin

Using Google’s Embedded AI Art Generator in Slides

I have signed up for the Google Workshop Labs because I am curious about how Google will be weaving its Artificial Intelligence tools into platforms like Google Docs, Slides, Sheets and more.

I noticed that the image generator tool appeared in my Slides the other day, so I took a poem from yesterday’s morning writing (via a one-word prompt off Mastodon – “specimen”) and tried it out with the five-line poem. I set “sketch” as style of art and then used lines of the poem as the text for image generation.

The slideshow poem is embedded above or you can view it full screen here.

I’m still not sure what to think of the visual results, but the tool is certainly handy, in one sense, and easily accessible, as it is located right in the Insert Image toolbar.

Google AI Art in Slides

Peace (in pencil),
Kevin

Generative AI Presentation 1: AI, Ethics and Education

Word Cloud: Teachers and AI

I visited our Western Massachusetts Writing Project Summer Institute yesterday to give a workshop presentation about Artificial Intelligence, Ethics and Education and we began with an interactive Word Cloud activity. The prompt was: what words or phrases come to mind when you think of Artificial Intelligence. The Word Cloud above is their collective responses.

You can see that most had more of a negative reaction than positive, and the conversations about reservations about AI in society at large, and in schools, more specifically, were enlightening.

Many had never even tried ChatGPT or knew about Google Bard, never mind the raft of other tools that have come out. Only one participant in a K-12 school had had any discussions at the school level about the emergence of Generative AI, and no one knew of any school policies that have been put into place.

Our ethical discussions revolved around issues of privacy, of copyright, of bias and more, and then we moved into the ideas of whether Generative AI belongs in classrooms, and if so, how and why. This led to conversations about leveling reading texts, engaging in AI “thinking partners” for inquiry, creating outlines on topics and more. We did talk about copy/paste plagiarism (one of the Summer Institute facilitators — and our new WMWP site director — is a university professor, and she noted this is a big discussion point in content-area departments more than the English department).

The big takeaway for me was a rich discussion about the authenticity of student writing, and how teachers must be attuned to the writers in front of them, and the possibilities of using technology to guide that process (or not) but not to replace that work. Tapping into the writing activities gives students a unique voice and authority that AI tools just don’t have. (Yet?)

I used Curipod as my presentation platform — it is another AI site for making presentations — and if you want to see the presentation (and make a copy of it for your own use), you can by using this link.

Peace (and Wonder),
Kevin

AI Analogy: Wedding Bands and Disc Jockeys

Drums
Drums flickr photo by JP Carrascal shared under a Creative Commons (BY-SA) license

My father, before he retired, was an accountant by day, drummer by night. When I was growing up, he was always busy with music. His Fridays, Saturdays and sometimes Sundays were spent keeping the beat in a variety of different wedding and dance bands, and he traveled all over, playing with different combinations of musicians every gig. (He also gave private drum lessons during the week, but that’s not really part of this story.)

And then, quite suddenly, the gigs began to dry up. It was the start of the DJ Revolution, where one person or two with a mixing board could replace an entire band of live musicians. For wedding and party planners, the cost differential of hiring a single DJ versus an entire band was huge, and for many, the finances was the decision factor. Understandable, but regrettable.

What was lost was not just my father’s regular gigs, which petered out over a stretch of time, but also the element of live music at these events, which I contend is a huge loss for making a wedding or party something special. Oh, I know DJs can do a fine job of finding the right songs and making the right mix to get our hands in the air and all that (one of my sons is now a part-time party DJ, ironically), but I’ll always argue that being in the presence of a good live band is a transformative experience. You feel the music being created, and the connection between audience and musician in the same room is an energy you can’t quite replicate with machine. Watching a human performance is a visceral experience, tying an event to memories in ways that digital tracks cannot.

Anyway, my father’s gigs dried up as DJs became more prominent, and I was thinking of this shift recently as the ethics of the data sets of Artificial Intelligence, and the ease of use to create art with Generative AI, comes more into focus. If AI can write a script, or make a piece of art, or compose a piece of music, or produce a video, or whatever, and the cost differential is the bottom line for companies that would have otherwise hired real people to make real art, then the world will shift.

Like it did for my father and his generation of musicians.

We see this same issue now playing out in the strikes in Hollywood with writers and now actors, who, among other things, worry about if their creative talents will be replaced by AI in the future as cost-cutting measures. It’s a legitimate concern, I think.

What will be lost when my father’s generation of live musicians faded into the shadows of the machine is similar, too: the authenticity of interactive experience, of being in the immediate presence of a piece of art that was conceived by the human mind and brought into being through the creative process for an audience of one (you, when you experience the art, in whatever form that art has become).

I’m not sure how I feel about this cultural transformation, or if it can be slowed or altered at this point in time. Probably not. But I already mourn the loss as a member of the audience and as a maker of art, you know?

Even as I play around with Generative AI platforms and use AI to figure out its potential (and maybe new art forms, which is potentially exciting), I know I’ll hang on to the pen scratches on paper, to the power of a musician’s solo that is different in every performance, to the synergy and energy of actors in a scene, and to much more, but will the generations after us even care whether the art is live or if it synthetic?

I hope so.

Peace (Pondering It),
Kevin

Google’s New AI Duet Workspace: Hummingbird Haiku

I saw an invite to my Google Account to pilot Google’s integration of AI into its various apps called, I think, Duet AI. They are rolling out a bit of it at a time, and after some hemming and hawing, I decided to give it a go (I might still opt out at some point but given all the inquiry and experimenting I have been doing already, it seemed like another step). Google has an explanatory page about Duet here.

I experimented with Duet in Google Docs after deciding to do some poem collaboration with the AI (which I assume is powered by Bard). I began with a haiku about hummingbirds, and then asked the Duet AI to write one, too. It did, and there are some options for tweaking the text, if wanted. You can Recreate (I didn’t find this worked all that well but maybe that is because the haiku poems were short) or Refine (with a few different parameters).

Then, I kept going, back and forth (another game of AI Chat Tennis), and I even tried to get it to turn the Document of poems into a Presentation (which I saw happen in a demo but it told me that feature had not yet rolled out).  The BOLD poems are mine and the ITALICS poems are via Duet.

Google AI 1

I even asked it to generate a list of possible titles for our collection. I had to recreate this a few times. Most were drab. One ended up OK.

Google AI 2

I later added the collection of poems to Giphy and added my own animated birds in motion, because … well, hummingbirds, right?

My initial reaction — I could see Duet being helpful as a thinking partner. The tool didn’t feel too intrusive (it sits off to the side with a little icon) and when I opened it up, it gave me some possible ideas for use with my writing. I haven’t looked at my email platform yet nor the presentation platform, but I will.

I do wonder about how and when and if Duet will be integrated into Google’s school platforms, and what that might look like. (I think Google has said they are NOT pushing Duet and AI into school Google networks … yet)

Anyone else given Duet a go?

Peace (Experimenting),
Kevin

Observations On A ChatGPT Collaboration

one path, diverged

AI Image by Dall-E

Sometimes, you just need to play to figure out something new.

Yesterday, a few friends joined me in exploring a collaboration feature within ChatGPT that allows you to share out a query/response, and another person can then build on it, and then share it back out.

It works OK, but there is no overview map of where the strands get taken, and by whom. It’s easy to be using a link that someone else has already used and advanced, and you’re still in the past. Things can get confusing, quickly. It would make more sense to do this with a single partner, sharing information and queries and responses back and forth (this could be a classroom activity between two students, perhaps?)

There were about a half dozen friends working at various times, sharing back links here at the blog or on Twitter, and two strands emerged at the end, with a slight diversion. I brought both to a “close” this morning by asking ChatGPT to write a poem in the style of either Emily Dickinson or Walt Whitman.

Strand One: https://chat.openai.com/share/c70b7769-729e-4e8e-83da-f1ee1974e53a

Strand Two: https://chat.openai.com/share/09319ab7-363b-4f64-b297-2a1a0b59fbbf

You can still play around with the strands, if you want, as this game of ChatGPT Tennis, as Wendy calls it, is everlasting (I think). Just share the link to your extension somewhere and we’ll play on.

Peace (and Play),
Kevin

Feature: Chat Collaboration (or ChatGPT Tennis)

chat collab screenshot
My CLMOOC friends, Sarah and Wendy, were doing some riffing off a ChatGPT response the other day to a DS106 Daily Create when I noticed they were using a feature I had not seen before, allowing you to “share” out a ChatGPT query and response via a link.

This feature allows other people to then access and use the original query/response for further prompting — essentially giving an opportunity for what Wendy delightfully called “ChatGPT Tennis” (as in, return the volley to someone else, and build on an idea, then send it back).

I am curious about the ways this might be used by partners or teams of people, to work an idea into either variations or perhaps to further hone in on a kernel of an idea together. Wendy and her thinking partner do a nice job going over some possibilities in her Elevate Postcard video.

I did ask ChatGPT to create a 25 word story (about thunderstorms) and it generated the link for sharing, so if you want to play along and see how it works, here is the link to the original response. You will need to share the links of any iterations, however, as the original owners don’t get any notifications (which is too bad). You could do that here, in the comments, or on Twitter or Mastodon. Or you can decide not to share back out. It’s up to you.

Peace (Volley For Serve),
Kevin

Dialogue Notebooks: Using And Creating AI Thinking Partners

NowComment and AI: Dialogue Journal

I am fortunate to be connected to folks in my National Writing Project circles like Paul Allison, who administers Youth Voices network and the NowComment annotation site. Paul has been at the forefront in my circles in thinking a lot and experimenting with the possibilities of Generative AI as a technology tool that could help readers and writers.

I had noticed that Paul and a few others NWP-affiliated folks were experimenting with the concept of AI Thinking Partners at NowComment — AI generated bots that engage with a text — but I have only just dipped my toes into what he has been doing, now that my school year is over. Then I saw him ask folks to experiment by keeping a daily Dialogue Journal in NowComment, and experiment with the AI Thinking Partners that he and others have been building out (using the API of ChatGPT).

I jumped in and I have made my journal (from June 24-July 1) public, so feel free to peruse my daily writing, which was themed each day on thinking about Artificial Intelligence (which I have been doing a lot since ETMOOC2 in the Spring). Then, after my journal writing, I would ask the Thinking Partner to engage a bit in discussions with me in the margins.

Read my journal

Paul also met with me to talk through some of the steps to creating my own Thinking Partners, and I made three: JazzHands, with a creative arts persona; HaikuHere, which is designed to turn a text into haiku; and WriteOutRanger, which is an experiment for the Write Out project to have an AI “park ranger” engage in discussions about place-based learning.

AI Thinking Partners

I am still thinking through the elements of all this, and wondering in the back of my head how this might be helpful or not to students in a classroom. For now, I am in experiment mode, playing around and trying to learn through experience the possibilities of embedded AI Thinking Partners. I know I need to learn a lot more about Prompt Engineering, and Paul and others have gathered together a lot of resources to explore on this topic.

Peace (and Inquiry),
Kevin

PS — Do you want to set up your own AI Dialogue Notebook in NowComment? Here are the instructions at the NowComment site. Let me know how it goes.