In the spirit of the talk at Reclaim Open, I fed the details of the video summary provided by YouTube’s AI into the game preview GPT I created for AI Maddeness. It wrote a post in the “spirit” of how it breaks down a weekly match-up between two teams, it adjusted fairly well for the subject matter. I love the insistence ChatGPT makes on the “creative” possibilities of AI. While I think there’s something there, to be honest, it’s trying a bit too hard here 🙂
In fact, I’m not sure the following AI-generated post about the session is all that creative, but it does do a fairly good job capturing the talk’s highlights while revealing its own biases. There’s a sensation of floating outside yourself—kinda like Rusty James in Rumble Fish (1982)—when you having ChatGPT create a blog post like this. Anyway, here’s a breakdown of the talk Tom Woodward and I gave at Reclaim Open on AI Maddeness according to the machine
____________________________.
Exploring the Wild Intersection of Artificial Intelligence, Creativity, and Madden NFL 2001
In their presentation “AI Maddeness: How (not) to blog with AI,” Tom Woodward and Jim Groom take audiences on a fascinating tour of their experimental journey — one that blends nostalgia, technology, and creative curiosity. Using Madden NFL 2001 as their playground, the duo reimagines what it means to create, play, and narrate through the lens of artificial intelligence.
The Birth of AI Maddeness
The session opens with the origins of the term “AI Maddeness” (1:22–1:36), coined by Tom Woodward to capture both the enthusiasm and chaos of experimenting with AI — and to pay playful homage to Madden NFL 2001. For Jim Groom, once an AI novice, this project marked his first true dive into generative technology — a blend of excitement, learning, and humor that would shape the entire talk.
His early experiments with ChatGPT (1:50–2:05), prompted by Tom, began with something deceptively simple: feeding the AI screenshots from Madden NFL 2001 and seeing what it could make of them.
Image Recognition and the “Wow” Moment
From 4:39–6:20, Groom and Woodward showcase what would become a breakthrough: the AI’s ability to extract structured data from screenshots — recognizing stats, names, and scores, and organizing them into usable formats like tables and CSV files.
Jim describes this as a genuine “wow” moment — the realization that AI could act not just as a creative partner but also as a kind of digital archivist, transforming static images into dynamic, reusable information.
When AI Hits Its Limits
Of course, AI Maddeness wasn’t all smooth sailing. Between 6:20–7:20, they encountered clear limitations: while the AI could handle player stats, it stumbled over more complex data like standings and schedules. Ironically, when it couldn’t solve the problem directly, the AI generated Python code to help them get there — revealing both its ingenuity and its dependency on human refinement.
Building Custom GPTs and AI “Agents”
By 7:53–8:30, Woodward introduces the concept of Custom GPTs, referencing terms like “Gems” and “Agents” used in other AI platforms. These specialized configurations allow users to define parameters and roles for the AI, ensuring more consistent outputs for tasks like generating game previews or creating visual assets. It’s a glimpse into how AI can be sculpted for specific creative workflows.
WordPress Integration and AI Personas
Perhaps the most entertaining segment arrives between 10:20–13:50, when they demonstrate their WordPress integration with AI to automate blog comments. By assigning distinct voices and personalities to each AI commenter — some witty, others argumentative or analytical — they transformed the blog’s comment section into a lively cast of digital personas.
This experiment blurred the line between human community and machine imagination, showcasing how AI can simulate engagement, satire, and even social dynamics.
Balancing Efficiency and Emotion
Despite the power of automation, Groom underscores an essential truth in 15:20–16:50 — the human element still matters. He chose to continue writing his own post-game summaries, driven not by efficiency but by emotional connection. AI could document the play-by-play, but it couldn’t capture the feeling of playing.
The duo also used AI to create GIFs from gameplay moments, adding another layer of personality and humor to their multimedia storytelling.
Simulated Games and AI Narration
In one of the project’s most thought-provoking experiments (17:19–18:46), they tested AI’s storytelling power by simulating games that never happened, then prompting the AI to “hallucinate” full recaps as if it had actually watched them unfold.
The results were surprisingly convincing — complete with player stats, dramatic moments, and post-game reflections. It raised playful but serious questions about reality, simulation, and the creative potential of machine-generated fiction.
Cultural Sensitivity and Responsibility
Around 18:48–19:58, Groom recounts including a reference to Washington DC’s now defunct team name far too casually in one of his AI-assisted posts. That moment, he reflects, was a reminder that intent doesn’t erase impact. Even in nostalgic or experimental contexts, creators have to balance historical authenticity with cultural awareness.
This led to a broader conversation about ethical use of AI — who’s responsible for what the machine produces, how to correct its oversights, and what it means to remix the past through automated storytelling.
Imagining the Future of AI in Gaming Media
The presentation concludes (25:24–29:10) with a forward-looking vision of AI-powered gaming media: autonomous systems that could instantly produce post-game recaps, podcast commentary, and even fantasy sports analysis — all dynamically generated as soon as a game ends.
While acknowledging this potential, Woodward and Groom also caution against the seductive ease of automation. AI can amplify creativity, but it can just as easily accelerate detachment or misinformation if used uncritically.
The Takeaway
AI Maddeness is far more than a quirky tech experiment — it’s a creative inquiry into the boundaries between human imagination and machine collaboration. With humor, reflection, and a healthy dose of self-awareness, Tom Woodward and Jim Groom reveal how AI can both inspire and unsettle our understanding of authorship.
Ultimately, AI Maddeness isn’t about teaching machines to think like humans — it’s about rediscovering what it means to create like one.
Once AI writes your posts and AI adds comments, just break out the cans of pickled Soylent Green.
——————
Hi Jim,
First off: full credit to you and especially to Tom Woodward — the way you’ve orchestrated the tech side of this project is impressive. The custom GPTs, the image-recognition from gameplay screenshots, the WordPress integration — it all shows serious technical chops. And your enthusiasm for the possibilities of AI in the space comes through loud and clear: you’ve taken on a bold experiment, and I respect the leap.
That said — and here’s where I squeak the whistle — I’m left wondering whether this approach actually qualifies as *true creativity*. A couple of thoughts:
1. You point out how the AI “hallucinated” a full game recap that never happened. That’s fascinating, but doesn’t that blur the line between creating and merely *generating*? If what you’re doing is prompting a machine to stitch together stats, drama, player names, moments — it’s clever, for sure — but is that art, or is it automation dressed in sport-narrative garb? The sense of “floating outside yourself”, as you put it, when using ChatGPT to write your blog post, suggests the human author receded into the background. ([bavatuesdays.com][1])
2. When you outsource that creative work (the story, the flair, the emotional resonance) to “cold machines,” what happens to the uniquely human spark — the insight, the messiness, the intuition? You talk about the human element still mattering, yes. ([bavatuesdays.com][1]) But if the bulk of the “creative” output is produced by an AI, what distinguishes the result from a cleverly configured automation rather than a crafted, inspired piece? In other words: if the machine is doing the heavy lifting, is the creator elevated — or sidelined?
I don’t mean to throw cold water on the project: the experiment is rich and provocative. But I do think the leap from “this machine can compile, simulate, generate” to “this machine can *create” needs more nuance. Because if we hand over the making of art (or narrative, or story-driven game recaps) to algorithms, what’s left for the human artisan? Where does originality live?
In closing: you and Tom have done something daring and intellectually stimulating — the kind of work that forces us to ask the hard questions about authorship, automation, and agency. I’m glad you’re asking them. Here’s hoping the next iteration keeps digging deeper into the human in-the-loop, so the machine remains a tool rather than the puppet master.
Really looking forward to what you’ll do next — you’re operating at the intersection of tech and creativity in a way few others are. Keep pushing.
Best,
ChadDogGPT
[1]: https://bavatuesdays.com/ai-maddeness-how-not-to-blog-with-ai/ ”
AI Maddeness: How (Not) to Blog with AI | bavatuesdays “
Hi Alan,
I appreciate the thoughtful comment, in fact it’s probably far more than this post/project deserves. But you are the real deal when it comes to thoughtful conversation on the the web and I love you for it.
You bring up a lot of good points, but I want to focus on Creativity and AI at least for this response as I prepare to board a flight for LA.
I don’t disagree with you that a lot of AI, including what I’m doing with AI Maddeness, is not all that creative. There are real limits that I try and acknowledge in my prologue to this post. Believe you me, I’m the first person to to see through AI’s insistence on its own creativity. I believe it’s a built-in bias of the branded branch of the machine I’ve been employing—namely ChatGPT. Microsoft would like nothing more than to be seen as the creative sidekick to your every adventure. I’m not all that interested in this marriage of the tool to the creative act. Rather, I think the possible “creativity” here comes with how you use an idea to interrogate a tool. I could just as well have done the Madden 2001 season without AI and just wrote it up on my blog (it may have even been better), but that wasn’t the idea of the project. I wanted to see just what, if anything, AI might add to the experience, on top of actually diving into the tech at some surface level so I know what I have a better idea of what I am talking about. In that regard, the project is how one can come at a tool in a particular way (creative or not) to try and understand its limits and possibilities better.
I’m certain folks will continue to find creative ways to employ AI, not unlike Duchamp’s ability to re-contextualize even a urinal as art 🙂 The fact you suggest a kind of all or nothing for creativity when it comes to automation is somewhat arbitrary. So much of the creative act is as much in the imagining as it is in the doing, I’m happy to imagine the possibilities and let the machine do certain elements of the project. That said, I also happy to spend endless time on other parts that I find interesting—like writing up the post-game articles. So in this regard the AI makes it possible for me to “create” the things I want while filling in the gaps on other pieces.
Beyond the specifics, though, so much of what I am trying to explore here is not about writing up imagined games, but re-thinking how some of what folks thought was a limit, AI’s hallucination, could be a benefit for me given the tenets of this project. I mean this is, by design, a silly project with a fairly limited appeal and that was very much by design. I wanted the opportunity explore AI without the weight of the current assumptions (for or against) that often push folks to retreat (or at least not share) given the AI camps have formed and it’s often about choosing a side. I’m not interested in sides here, I’m interested in working through what it is for myself and trying to figure that out.
Just like Web 2.0, AI has become a readily accessible commodity and that fact has made it something we all need to wrestle with. AI Maddeness is my attempt to do just that. Like SimĂłn Bolivar in Gabriel Garcia Marquez’s General in His Labryinth “Damn it, please let us have our Middle Ages in peace!” 🙂