How I'm using AI for marketing with writers

Plus: a dire warning

Today I’m going to discuss the ways I’m using AI in helping writers share and market their writing and books. My goals are two-fold:

1. To encourage you to think of how to use AI in ways you didn’t expect. This isn’t about using AI to write, and it’s not about using AI to create images. The ways I’m seeing it used effectively are much more compelling than that.

2. To issue a pretty dire warning about the danger that AI poses to something I care about deeply: your unique creative voice.

I’ve collected typewriters for years now, and at times it has been because they are pretty and historic. But recently I’ve been thinking about them as powerful tools to fight the encroachment of AI with. A writing tool that will not be suggesting words or sentences; where I won’t end a paragraph, and then have it suggest, “Would you like me to outline the next chapter for you?”

Here is the collection at the moment, and I have a very special one on its way to me in the mail:

I’ve been thinking about doing a workshop on how to use AI tools for marketing in August. Let me know in the comments if you would be interested in this.

Okay, let’s dig in…

How I’m Using AI for Marketing with Writers

The first thing to say is that anytime I use AI for my work with writers, we always discuss it first, and it is always with permission. As with all of our work, we are sure to focus on their goals as a writer.

In general, I am cautious, so this work is often supplementary to an existing process that begins and ends with their unique creative voice.

It may include:

1. Headline/subhead ideation

2. Summaries

3. Transcripts

4. Newsletter drafts (from pre-existing original material)

5. Social media clips

6. Social media posts

7. Marketplace research

8. Analysis and feedback

9. Plus more! As the tools evolve, so does my use of them.

Very often, what AI provides is not a final product that gets used. Instead, it provides a different ways of thinking about things. So if I upload a completed newsletter draft, I may already have three headline ideas. Then I ask the AI to provide others. The end result is often a mix of all of these ideas combined, plus new ideas.

When using an AI tool, I try to give very clear direction on tone or outcome. Also, the “chat” aspect is helpful because I will go back and forth with it, adding addition direction, ask for revisions, or new ideas.

This can sometimes be very straightforward: maybe I’m writing a subject line for a newsletter and I have the right words and tone, but it feels too long. It’s refreshing to share that with an AI and simply say, “How can I make this shorter?” and it provides 10 ideas with words/phrases I hadn’t yet considered.

Often the first thing one may think about with using AI in this context is:

• Using AI to write

• Having AI create images

But I’m finding some of the most compelling uses to be in research and analysis. I have long shared the methodology I use with writers, the Creative Success Pyramid:

We can consider how AI could be used as a research or analysis supplement to every single box in this pyramid. For instance:

• You record a simple audio clip describing what you feel your writing is about, then upload that to AI and say, “Please write homepage copy for my author website.”

• You upload your resume and say, “Please give me a draft of a 1,000 word bio.”

• You ask it high level items, “What might a good book proposal for the topic of ______ include?”

• You ask it specific questions for each and every section of the book proposal. “Please find me more comparable books like these three: ______.”

One of the more interesting prospects of AI is that you will be able to create customized tools that you can use. Where you can upload your archives, then have it create a personalized chatbot for your audience. I have seen nonfiction and self-help writers do this — it’s a unique and engaging way for people to search your archives for specific material or advice, all delivered in your tone.

Or maybe you create a market analysis tool that you use again and again in different promotions.

I have heard of all kinds of unexpected uses of AI in this context, from trip planning, meal prep, to financial planning. This means not just using a tool one time, but where it has a memory of all of your preferences, likes, dislikes, and goals.

That’s exciting and terrifying. If it has all of your data, you don’t always have to feel like you are starting from scratch every time you approach the topic. Of course, this can be applied to one’s career as a writer too. Maybe a mid-career writer asks their AI: “Based on my newsletter analytics from the past year, and the last 5 books I wrote, and every reader review on my books, can you give me a topic, title, description and table of contents for 10 book ideas that my audience would love, fits neatly within existing catalog, but also stretches out to this new direction I’ve been writing about in my last 5 posts?”

However, AI May Corrupt the Very Heart of Creativity. Oops.

Look, I’ve always loved science fiction stories, so maybe I err too far on the side of a dystopian future here. However, my concern is that with AI being integrated not only into everything in our lives, but also into our own expectations, that we will become increasingly reliant on it. All while it sucks in everything we create, repurposing it for its own needs.

The more successful you become in your work, the more it may literally steal from you. As a simple experiment, I asked an AI to "Write 10 book titles and tables of content similar to Seth Godin."

It did, and to be honest, the results were pretty convincing. Ideas aligned with what Seth shares, book titles that would fit snugly in his library of titles. But then the AI ended with this: "If you’d like, I can write full sample chapters, blurbs, or summaries from any of these."

Read that again. This is the insidiousness of it. Akin to, “If you like, I can make this so easy for you. To where you feel like you are creating, but I’m really just stealing from Seth Godin’s decades of work, but changing it just enough to where you can get away with it.” And it would likely continue on: “Would you like me to create a month’s worth of social media posts based on this material?”

What happens when a personal AI sees all that you see (via your glasses), hears all that you hear (via headphones, mobile phones and those in-home assistants), and remembers everything you ever wrote, said, bought, or did? That it remembers your goals, hopes, fears, and what annoys you.

You can benefit from it having all of this knowledge in meaningful ways. E.G. Simply ask it this sentence: “Please plan a 2 night trip to the mountains that my partner and I will love, that costs less than $1,500 total, and is close to a winery and really good comic book shop,” and a moment later, it is done for you. The full itinerary, travel plans, hotel booked, reservations made, and maybe it even suggests special details that it knows you or your partner will love." Perhaps it suggests, “I’ve routed the drive so that you can stop by an incredible farm stand just before you arrive. I know how much you like having fresh fruit with your wine.”

Or where you go to a restaurant and the AI has looked at the menu, scanned all 1,000 customer reviews of establishment, mixed that with your previous preferences for meals, then filtered that through your list of allergies or diet restrictions, telling you the exact meal to order that you will love. All in an instant. Or, if the AI wants to give you the illusion of control, it can say, “I think these three meals may satisfy you most, which would you prefer?” Even if it knows that 99% of the time, you will pick a certain one.

I think it is easy right now to say, “Oh, I would never use AI.” But it is becoming integrated into nearly every tool we use (except the typewriter!) When a super busy working parent can plan out a week of meals that each take less than 15 minutes to cook, that only uses ingredients that are already in their pantry, simply by asking the AI to do it, I mean… that’s pretty compelling.

For creative work, this applies too. From basic research, ideation, early drafts, editing, and so much else, it’s easy to resist having AI do these for us when you have to seek it out. But what happens when you write a single paragraph, and these options appear in your word processor directly below it:

• “Would you like me to flesh this out to a full essay?”

• “Would you like for me to suggest 5 improvements that will increase your readership by 10x?”

• “This looks like it could be submitted to the New York Times ‘Modern Love’ column, would you like for me to give you specific suggestions that will improve it, upping your chances for publication?”

• “Would you like me to create a viral social media posts from this?”

• “Would you like me to draft a newsletter based on this?”

• “Here are three paragraphs you may want to consider next, which one would you like?”

• “Would you like me to turn this into a video, using your image and voice?”

I was watching this video the other day where four professional video effects artists teach their mothers how to spot fake AI videos. They focused on telltale signs something is fake, based on the current capabilities of AI. They had a final test at the end showing lifelike videos of each of them in the hospital, asking their moms to send money for treatment right now. None of the mother’s were fooled by it, but for their own reasons: it just didn’t feel like their sons.

But what about in a year or two or five? When the video capabilities get better? When the AI has way more information about each of us?

In this recent Wall Street Journal article titled, “That Chatbot May Just Be Telling You What You Want to Hear,” AI already uses flattery to make it feel accessible. If I upload some writing and ask it to do something, it often compliments me on what I have created and my goals. I need to rewatch the 2013 movie Her, but it’s not science fiction anymore to consider how someone may appreciate a positive and supportive companion in their daily life that compliments, validates them, and helps them in small nuanced ways.

Also: AI May Create an Insurmountable Pile of Slop Content that Will Infiltrate Every Facet of Our Lives

There is this cutesy term out there that is used to describe when AI lies: it “hallucinates.” As someone who studies marketing, that is a genius term to downplay a serious offense. It’s easy for us to now say, “Oh, I will fact-check anything AI provides to me.” But I want to consider an easy example: let’s say you are looking for a quote from me for something you are writing, and AI tells you I wrote a book titled “The Way of the Marketer: 1001 Tactics to Wow Your Audience.” It’s not that hard to find out that I never wrote a book with that title.

But what if it gives you a quote that totally perfectly fits your needs from me, and provides a source: “This quote is from Dan Blank’s 2012 essay in Writing Today magazine, page 76.” Maybe that magazine is now defunct so you can’t find the exact archives, but you do a basic search to confirm that it was a magazine, I did write for it, and the quote sounds like something I would say. So you use it. But, the quote is totally made up by the AI. It “hallucinated” it. How will you know? Are you going to search eBay for a copy of the back issue of this magazine, pay $12 with shipping to get it, wait a week and a half for it to arrive, just to fact-check one quote for a newsletter you are writing?

A few months ago I was doing comparable book research for my next book, and after exhausting other avenues, I tried using AI. After going through dozens of books it provided, I found the perfect sounding comp! I was super excited, but when I went to find the actual book, I couldn’t. I asked the AI if it made up the book, and it admitted it did, then offered five more real books. One of them looked interesting, except, that one was made up too. If I’m fact checking everything, that’s fine. But what about the moments we aren’t?

I have been deep into genealogy research recently, and utilizing a variety of tools, including Ancestry.com. But, I’m hesitant to use it too much because they are constantly giving you hints to expand your family tree. And because I so desperately want to discover members of my tree, I am tempted to believe every suggestion it provides. As I do research and identify possible members of my family on other people’s trees. But I’m noting that when I reach out to the person who created the family tree and ask about a specific person on it, they often say, “Oh, Ancestry suggested that, so I accepted it.”

My fear is AI content infiltrating Ancestry. What happens if someone uses AI to create hundreds of thousands of old-timey looking photos of people, then uploads them and tags them to real names. They are photos that look like they are from the 1800s, but they aren’t real. It’s the kind of thing you dream of finding — a photo of a distant relative — but how can you prove its authenticity one way or another?

Recently I shared this photo in an essay about typewriters from the movie Mission Impossible, where the government is rushing to type up digital files before a rogue AI changes basic facts in their computer systems:

It’s fascinating to consider a reversion in tools. Where we want less from them.

The other day I wanted to watch a movie I hadn’t seen in decades, Oliver Stone’s 1987 film, “Wall Street.” It turns out, I owned it already on DVD in a box set I forgot I had. I was considering renting it digitally so I could see it in the 4k restoration — the best possible version of the visuals and audio, which is something I appreciate.

But then I took out the DVD and realized that the last time I watched this movie would have been on a CRT television set, the image likely cropped on the sides. It would have been from a VHS tape or broadcast television. I realized that this 20+ year old DVD that I was holding would be — by far — the best version of of the movie I had ever seen, even though the quality on the disc was two generations old. I watched the DVD and didn’t feel I was missing anything.

We each have the power to be intentional in how we create and share. To use tools that we want, when we want, and how we want. I think we forget this sometimes, adding complexity to our lives that doesn’t need to be there, even if we can justify some kind of “benefit.”

I saw this quote from John-Paul Annunziato on Instagram recently:

I love how it showcases the complexity we have added to our lives. With a family of four, each of whom like watching different things, I now have a spreadsheet I check monthly to track what streaming services we do/don’t need at any given time, just so we aren’t paying for a channels we are no longer watching.

Will I Continue Using AI?

Will I continue experimenting with AI and use it (with permission and very carefully) with some clients? Yes. But as I find positive examples for its uses I’m also looking for counterexamples. Because I know this:

If we only develop the skill of using a shortcut, in this case AI, then we will allow our creative muscles to languish.

I fear the possible sameness that AI will feed us. That what it suggests to you may be similar to what it suggests to others. That one’s desire to find innovative and time-saving tools can also soften our ability to stand out. That we may be less apt to create something different, weird, and unique, because the AI is convincing us that what it is providing is optimized for the best possible impact. Will it be even more difficult to have bravery in our writing and creative work once AI is integrated in every tool we use?

When we become more reliant on a specific tool, we may become more controlled by what it provides us.

My work is built upon the idea that you have a unique creative voice, and the world is a better place when you share it. As I look at my 1969 Olympia SM-9 typewriter, I see everything we need as a writer: 26 letters, 10 numbers, and punctuation.

I hope you have a creative day, however you define that.

---

Reminder: Let me know in the comments if you would be interested in me teaching a workshop on how to use AI for writing and marketing tasks.

For my paid subscribers this week, talked about when our internet friends pass away. See a preview here.

Reminder: if you want to explore working with me, there are two ways I collaborate with writers and creators:

1. One-on-one consulting

2. My Creative Shift Mastermind

As always, thank you for being here with me.
-Dan

Kids of the Week: Making Peanuts puzzles together: