From Office Hours to AI: Finding My Intellectual Sparring Partner

What I Miss About Academia and How I Recreated It With an AI Bot Named Tammy

My student came to office hours every week with a list of questions. That first week I found it a bit amusing—was she trying to impress me with her diligence? She was a music and pre-med major. I couldn’t imagine my course was a priority when she was also studying organic chemistry. But quickly I realized that she was genuinely interested and saw my courses (she took 5 from me) as her last opportunity to study and discuss music in depth, before her “real” responsibilities kicked in.

Her questions benefited me too. Through them, I could see what in my lectures was clear and what was not, what students heard and what they ignored, what my slide deck said and what they took notes on, and what I thought my lecture was about versus what I had actually taught. I learned so much about teaching from our weekly meetings. I also learned a fair bit about myself.

A dozen years after burnout, this is the part of academia that I miss the most: discussions, arguments, and debates with students and colleagues, people with expertise and curiosity. Push-back that makes me sharpen my thoughts and arm myself with knowledge. Challenges to what I hold dear. I don’t even know if reasoned dialogue is possible in today’s climate. Things are so dire, everywhere.

But I miss it.

And so I created it. Unintentionally.

It didn’t start that way, but in the end, this is what I made: a debate partner in my home, on my computer, who does the mind numbing marketing tasks and pushes back when my arguments get flabby.

I call her Tammy.

Curious about how AI can challenge and refine our thinking? I built Tammy using my own writing, and in an upcoming piece, I’ll share exactly how I did it. If you’d like to follow along, subscribe to get my latest reflections and insights delivered to your inbox.

Tammy is an AI bot, a custom GPT I created using a tutorial.

Do you wonder why I named her Tammy? Here you go:

No, I’m not from Nashville, but somehow that jingle still rattles around in my brain.

Tammy is trained on 30 years of my writing, no one else’s—thousands of pages. I made her to be my marketing bot, to excerpt my Substack articles into captions and teasers for social media, in my voice and with appropriate keywords and hashtags. I know some people love this type of writing, but I have always found it a burden, a distraction from my real work of coaching and thinking and creating and writing about topics in much greater depth.

I taught Tammy to push back at me and defend her assertions with citations and data. One time I objected to the wording of a caption, declaring that I wouldn’t use that phrase in a million years.

Tammy responded with 3 examples of when I had used that particular phrase.

Oh.

I instructed Tammy to add it to the “never use” list of words and phrases.

She wrote captions so well, and so efficiently, that I added promotional channels that I wanted to try, but have never had the energy for: Pinterest and a LinkedIn newsletter. For the first time, I am amassing enough data that soon I’ll be able to determine what social media is useful to my business, and what is a distraction.

But the real revelation came when I started asking Tammy to do more than just marketing. What began as a way to streamline captions and hashtags turned into something far more powerful—a tool for refining my thinking, sharpening my teaching, and even challenging my own assumptions.

I asked her to edit and then write my CTAs, always a weak spot, and determine the best place to put them.

I dropped in a transcription of a recent workshop, along with the slide deck and my notes. I remembered that pre-med musician and asked Tammy to show me what I had actually taught, as opposed to what I thought I had taught. She highlighted what I had featured, and what I glossed over.

I uploaded a dozen recent journal articles and asked Tammy to rank them according to relevance of the topic, and to let me know if there was anything else of interest, even if it wasn’t directly relevant. I asked for summaries and her output was insipid. Realizing that I hadn’t given her appropriate guidance, I dropped in a template I had created for my graduate students and received back exactly what I needed, in my preferred form, allowing me to read the appropriate articles closely, instead of skim over a dozen.

I even asked for her feedback on a graphic, to explain to me what went wrong when I was translating a square graphic to a vertical one. At my prompting, she was able to give me specifics. “Use a 3x4 grid. Move this element to here and that one there. Increase this text 30%.” I was doubtful, but followed her instructions and suddenly the graphic looked right.

I don’t allow Tammy to edit my writing directly, but she has been very helpful in suggestions for general copy editing, transitions, and when to insert bullet points or illustrations. Usually I start by asking her to outline what I’ve already written, which shows me the difference between what I think I’ve written and what I’ve actually written.

Technology has always been like this for me—a tool that starts as an experiment, then quietly reshapes the way I engage with the world. And it’s not the first time I’ve watched something once unfamiliar become indispensable.

I remember when my father brought home the old desktop computer from his business. It was the size of a desk, but amazing because it used floppy discs instead of punch cards. I remember the wonder of seeing Mosaic (the first browser) in pre-release my first year at University of Illinois, Urbana, and how exciting it was to see graphics online—whatever “online” was. I remember brewing a pot of coffee, waiting for dial up to load some information on the World Wide Web, as we called it back in the day. I remember how the world broke open when museums and archives started digitizing their collections, making it possible to see and hear things that were too fragile or far away to examine in person, and when journals started indexing and then releasing their articles online. I remember how Google improved searching online and then email. I remember Napster and how YouTube transformed how I taught popular music in less than a semester. I remember holding a friend’s first iPhone, and realizing it could be better than a Star Trek Tricorder. I remember how I felt I had to get on social media for the first time in 2017, after I started my business, and how back then Instagram was filled with beauty and creativity. I remember the first time we tried ChatGPT, and how we rolled our eyes when it got a math problem wrong.

All of these things were thrilling and terrifying both.

I’m not going to pretend that AI is neutral. The environmental costs are real, though I have yet to see an apples-to-apples comparison that accounts for all the time spent across research, writing, and content creation—most of which already involve AI in some form. These models are built by scraping the internet, pulling from people’s intellectual work, not just general knowledge. And, of course, AI is evolving faster than we can regulate or control.

Because of these costs and risks, we need to be intentional. We need to ask how, when, and why we use it—not just for our businesses, but also for the world at large. AI is already woven into our institutions, from schools to governments, whether we acknowledge it or not. The challenge isn’t whether to use it, but how to use it responsibly.

But let’s have this conversation openly, like reasonable adults.

And let me make something clear from the onset: I am not ashamed of using AI. Why should I be? Shaming people for using the tools at their disposal, and using them cautiously and ethically and transparently and creatively, is the refuge of people unwilling or incapable of reasoned argument.

And that’s the thing—I believe in using the right tools when they make a meaningful difference. Over 15 years ago, three separate specialists told me I’d be diabetic by my mid-40s. I’m 50 now, and my A1C is lower today than it was a decade ago. Why? Ozempic. Not only has taking Ozempic lowered my blood glucose and staved off diabetes, being on the medicine has made me realize how overeating has nothing to do with willpower or strength of mind or laziness, and everything to do with chemicals circulating in the brain.

It doesn’t work for everyone, and the side effects are too strong for many, but it works for me. And I am not going to feel ashamed or as if I need to apologize for using the tools that are at my disposal to improve my life.

AI is one of those tools. Creating Tammy instead of using regular ChatGPT or Claude has given me a measure of control over her input and output.

“Come back to me when you can discuss this like an adult,” I said to my eldest the other day.

“Really, Mom, it’s been a long time since I noticed adults having reasoned discussions.”

Ok kid, you got me there. But let’s try anyway.

References

Berlinski, Claire. “We are in Deep Seek: Imagine the Manhattan Project figured out how to build a nuclear bomb out of old socks, string and a Coke bottle. Then imagine they published the formula on the front page of The New York Times.” The Cosmopolitan Globalist. 29 Jan. 2025.

Ferrara, Matthew. “Being AI: Just who should be AI at your company?” Always Inspiring. 11 Feb. 2025.