How ChatGPT Summarized My Writing

Why AI Makes Me Uneasy

Let me preface this post by saying that the push by Tech Bros to insert generative AI into everything makes me deeply uneasy for several reasons.

The hype they’ve built in their rush to create artificial general intelligence (AGI) has steered eye-watering amounts of money away from investments in other areas. While I understand that investors in AI aren’t likely to put their money into solving the climate crisis or world hunger or gun violence or saving democracy (all much more worthy causes than enriching Tech Bros), surely there are other companies or other technologies that would give them a greater return on investment that are morally more supportive of society and the environment.

Using generative AI requires tremendous amounts of electrical power, straining our current energy grid. This is causing AI companies to invest in building facilities to create even more power, thus straining earth’s resources even further. (To reduce the unnecessary use of electricity by AI, I have disabled the AI feature on my browser.)

In training the large language models and other models behind AI, Tech Bros have run roughshod over the copyright of creators, knowingly using copyrighted works without permission. It’s called stealing, and I don’t take kindly to thieves. This is personal for me because my blogs have been used to train AI. While the New York Times and big-name creators have the resources to file lawsuits against giant tech companies, we small-time creators are out of luck.

That generative AI has been foisted upon us through building it into our browsers and other tech tools without any discussion about whether we want it or its potential negative effects (from its wrong answers to its ability to coach someone into suicide) frosts me. I’m not sure how we’d have an effective, society-wide conversation on new tech, but surely there is a better way than having everyone scramble to make sense of it and deal with the disruption.

Poking the AI Bear

That said, I have used generative AI to see how it works. Typically, I’m testing its boundaries to determine how much I can trust the output. I think of it as poking the AI bear.

As it turns out, I’m pretty good at writing prompts, which I attribute to my years of history research and helping others with their research. When you’re trying to help someone figure out what they are looking for within history resources, you’ve got to know what questions will help them narrow their search.

I also use my history background to poke the AI bear by asking it to provide me answers to and summaries of specific local history topics that are available online. Having spent 3 years writing a book on the history of Nathan Richardson, I use him to see what AI knows. It often gets things wrong on niche topics like this.

If you’re just getting started with generative AI, I suggest you poke the AI bear in this way – by asking it about subjects you know well, so you can determine when its responses are bullshit. This will hopefully instill in you a sense that you can’t completely trust AI’s output in areas you know nothing about. If it can’t get the stuff you know right, why would it get the stuff you don’t know right?

Using AI to Analyze My Writing

One thing I know well is my own writing, so I decided to ask ChatGPT to summarize the major themes of several of my blogs and analyze my writing style.

Here is the prompt I provided:

Read through the text below from the About pages on 3 of my websites and provide an analysis of my writing style and the major topics I write about. The sites are Without Obligation, The Woo Woo Teacup Journal, and Ask Little Falls. (I provided the text from the About pages in the prompt.)

Here is ChatGPT’s response:

1. History and Public History
   • You focus extensively on local history, particularly related to Morrison County, Minnesota. You aim to make history accessible and emphasize the importance of thinking like a historian.
2. Art and Creativity
   • Your writing often explores various art forms, particularly fiber arts, and your background in visual arts. This includes discussing your artistic family heritage and personal artistic endeavors.
3. Community Engagement
   • In “Ask Little Falls,” you address community issues and encourage civic engagement. You aim to foster dialogue among residents to improve community life.
4. Technology and Blogging
   • You discuss your experiences with blogging and technology, highlighting your commitment to privacy and open-source software. The concept of “blogging without obligation” is central, reflecting your flexible approach to content creation.
5. Writing and Authorship
   • You share insights into your writing process and projects, which include both fiction and non-fiction work. Your varied writing forms include articles, short stories, and essays.

• Freedom and Flexibility: The theme of exploring different subjects without constraints, highlighted in “Without Obligation.”
• Civic Responsibility: Encouraging active participation in community discussions and problem-solving.
• Personal Narratives: Infusing personal experiences and reflections into your writing to connect with readers.
• Innovation: Utilizing new media and technology to share ideas and foster communication.

• Conversational and Reflective: Your writing is personal and reflective, often speaking directly to your audience in an inviting manner.
• Engaging and Thoughtful: You provoke thought and encourage discussion, especially on community matters.
• Eclectic and Varied: Embracing a wide range of interests and subjects keeps your content dynamic and diverse.

For as cranky as generative AI makes me, I was both impressed and surprised by this response. I am impressed by how well it summarized the major themes of my blogs and the terms it used to do so: Freedom and Flexibility, Civic Responsibility, Personal Narratives, and Innovation.

Freedom and Flexibility sums up my approach to blogging over the past 19 years. I’ve never bought into the idea that a blog has to stick to one narrow topic for years and years, though I have done time-limited project blogs on a particular topic. My longest lasting blogs have all been free-wheeling in the topics I cover.

Civic Responsibility was a major surprise for me because I never thought of my writing as having this theme. Looking back at everything I’ve written and how often I discuss community and government, this should have been obvious, but it wasn’t to me.

I think ChatGPT also did a good job of capturing my writing style. “Eclectic and Varied” reiterates “Freedom and Flexibility,” so my writing style supports the variety of topics I cover.

When writing, I try to be conversational, engaging, and thoughtful, so it’s interesting that ChatGPT finds me successful at that. I hope my human readers sense that, as well.

Generative AI looks like it’s here to stay, which means we all have to figure out how much we want to engage with it. Poking the AI bear will help us determine where it can be helpful and when it’s responses are best ignored. We need to always be questioning this tool and not turn off our brains when using it.

---

How are you feeling about generative AI? What are your experiences with it? Do you find it useful, or can you do without it?