Learning AI Out Loud

Short weekly notes on what I’m learning, trying, questioning, and using as I figure out how to use AI in real work

Latest Column:
I’m Making Learning AI Out Loud More Useful

Week 11 – August 31, 2026

This week, I write about my plans for this column. I want to make it more useful for you. Please send me your questions about AI, content, writing, and content engineering.

I Thought I Was Keeping Up With AI. I Wasn’t.

Week 1—May 5, 2026

This is the first in a weekly series about how I’m using AI, what I’m learning, people I’m learning from, questions I have, and where I’m going next with AI.

I’ve been floating around Every.to for a few weeks now, reading their free newsletters and articles but blocked by the paywall. Was it worth it?

Should I give this company my hard-earned Canadian dollars?

As I dithered over the paid subscription, I became addicted to reading Katie’s Every articles. I follow her on LinkedIn. I sent her a connection request. Ignored—I get it, she’s very busy these days. Subscribed to her Substack.

And then I saw the Writing Camp. I really, really wish I had known about this sooner. Of course I tried to watch it—nope, subscribers only.

But still, I sat on signing up. Surely I could keep up and learn what I needed about AI on my own. And as for an AI writing tool? Yeah, no. Hard stop. 

I’ve got a writing tool: my big brain, 10 fingers, Microsoft Word/Google Docs, and a keyboard. I don’t need a tool taking over my words and sentences. I don’t want edits while I’m writing. I’m a professional writer after all.

A writer’s gotta write. It should be hard and slow sometimes. Right? Isn’t that the ethos of being a writer?

And then I started seeing posts about Claude Skills, content engineering, marketing engineering, and I reread Writing With AI is Harder Than You Think, How To Build an AI Style Guide, and AI Phobia Is Just Fear that ‘Easier’ Equals ‘Cheating’.

A switch flipped. I need to catch up. I’ve been using ChatGPT for a while now, I’m a proud owner of a paid account and I’m pretty good at prompting.

But I needed to be honest with myself and finally admit that what I’m doing isn’t enough anymore.

My husband likes to talk about the five levels of AI, and I was firmly stuck in level 1.5. I recently started using Gemini Gems with one of my clients to “automate” a lot of the heavy lifting that goes into writing and to handle the time-consuming, nit-picky work that always takes longer than I expect.

But this only gets me maybe to level 2. Not good enough. 

I need to know more. What are Claude Skills and how should I be using them? What is the difference between a ChatGPT Project and a Custom GPT (I asked ChatGPT—they don’t really seem to know either)?

How can I stay relevant in my marketing and content writing role? 

It comes down to being scared of being left behind. I like words. I’m lucky to make my living off of them. And I don’t want to be pushed to the curb because I haven’t kept up. 

So this is why I spent $390 on an Every subscription and why I’m giving up my free time to learn as much as I can.

I know it’s not going to be easy. I’m going to have to keep my brain—and my mind—open.

I’ll need to accept that there are better ways than what I’ve been doing. I’ll need to admit when I don’t understand.

I’m prepared to fail.

I plan to write weekly about how all this is going. Follow along for the honest account of what it’s like to be a 54-year-old writer trying to catch up and stay on top of all things AI.

Hopefully, my experiences will nudge you toward taking the AI leap—or point you toward new tools, people, and processes that can help you do your job better. 

Today I started the Airops Content Engineering Cohort and I watched the Every Writing Camp replay. In short: mind blown. And I’m feeling like I can get this. I can do this. We can do this.

How to Write with AI Without Losing Your Voice

Week 2—May 13, 2026

TL;DR

  • Writing well with AI still demands human involvement
  • AI doesn’t turn weak writing into great writing. It reflects the thinking, context, and direction you give it
  • Content engineering does not replace writers, marketers, strategists, or editors. It supports them
  • To build AI systems that actually help, you need to understand how you already work: your process, standards, audience, goals, and judgment
  • Context is everything. If you don’t know the who, what, why, when, where, and how behind your content, AI won’t solve the problem. It may make the problem bigger
  • There is a lot to learn!

AI can help you write faster, but it can’t replace the thinking behind good writing.

What a week it’s been.

I’ve been fully immersed in content engineering, writing with AI, and learning how AI tools find, cite, and use content. I’ve also started learning how to create Claude Skills, which has pushed me to think much more carefully about my own writing process.

How to Write With AI

First off, there is the right way and the wrong way to write with AI. 

We’ve all experienced the wrong way: word salads, if not x then y, sentences that don’t say anything, and long words that overcomplicate. 

Essentially, the type of content no one wants to read.

The right way to write with AI demands human involvement. 

AI is the helper. It can remove some of the friction around writing, but it does not replace the thinking, judgment, or decision-making.

Efficient writers are using tools like Claude Skills, custom GPTs, Agents, Spiral, and AirOps to help them outline, brainstorm, question, plan, draft, and edit content.

The process now looks more like this:

  1. Break down what it takes to create the piece of content.
  2. Map the steps: research, outline, strategy, CTA, keywords, thesis, audience, personas, examples, and editorial standards.
  3. Use those details to create Skills, prompts, or repeatable workflows.
  4. Give the AI tool supporting context, such as brand guidelines, writing examples, category narratives, and source material.
  5. Tell the tool what I want it to help me do.
  6. Then I review, question, refine, and decide when the output is good enough.

For example, I might ask the AI to use my outline Skill to create a blog outline. The Skill asks for specific information: thesis, audience, research, data, length, and purpose. Then it creates a structured first draft of the outline.

Or I might ask Claude to use my social media Skill to draft LinkedIn posts. The Skill asks about audience, post type, tone, CTA, length, and purpose. Then it creates three to five options for me to review.

The trick here is: the output is only as good as the input. The better the information I give the AI, the better the end result. 

And at every stage, I decide when the output is done. 

I’m driving the process. 

There is no copy-and-paste from ChatGPT or Claude into WordPress or Google Docs. 

I think about it like this.

Back in 1995 or so, I learned HTML and hardcoded websites. Cool skill at the time.

But that skill is not nearly as useful now. What matters more is that I’ve learned how to use WordPress to create better-looking websites faster.

The underlying skill did not disappear. It changed.

What I’ve Learned About Content Engineering

Content engineering helps teams build the systems, prompts, examples, workflows, and editorial guardrails they need to use AI tools more effectively.

It does not replace writers, marketers, strategists, or editors. 

It supports them.

The goal is not more generic content, faster. The goal is clearer, more consistent, more useful content with human judgment still at the centre.

The AirOps AI Marketing Cohort is focused on teaching us how to think and work like content engineers. 

The first four sessions have been incredibly useful. So far, we have learned:

  • How AI search works, including how tools cite content and why information gain matters
  • Why brand drift happens and how to control it
  • How to train AI tools on who you are, what you do, what you believe, what you want, and how you plan to use the outputs
  • How to use AirOps, especially how to create Playbooks
  • Why it’s important to study your own content process so you can turn it into Skills, prompts, and repeatable workflows that actually help

Essentially, it comes back to the same principle as good prompting: the better the context, examples, guardrails, and role definition, the better the output.

Junk in = junk out.

I spent over four hours creating brand guidelines, category narratives, personas, and context types. 

Now that this work is done, I can use it to build a system that helps me use AI more efficiently.

The goal is not to pump out more content. The goal is to create useful content that sounds like me, reflects my standards, and speaks to the people I want to reach about topics they care about.

Helpful AI Resources

Here’s a list of websites and people I’m learning from right now:

I’m halfway through this NPR Fresh Air podcast: Inside a Journalist’s Year of Using AI for (Almost) Everything.

AI Turning Point

Whatever tool you use, context is key.

To create better content and better systems, you need to give the tool real information: who you are, what you do, what you believe, what you like and dislike, what you stand for, and examples of your work.

You also need to fully understand why you’re creating the content. 

What gaps are you filling?

How are AI tools citing your brand right now?

What content do you need to improve your visibility in AI-generated answers?

Does your page structure need to be clearer, more useful, or easier for search and AI tools to understand?

If these questions are hard to answer, that’s probably the place to start.

Because the first step isn’t choosing the perfect AI tool. It’s understanding your process, your standards, your audience, and the kind of work you actually want AI to support.

That’s what I’m working through right now.

Each week, I’ll share what I’m learning and how it’s shaping the way I think about my work, my skills, and my future as a writer and marketer.

What If I Could Build That With AI?

Week 3—May 20, 2026

TL;DR

  • Learning AI takes play, curiosity, patience, and a dose of frustration
  • The better you understand your own process, the better AI can help you
  • If you want AI to build something useful, you need to know what you’re asking it to do.
  • You can’t shortcut the thinking

AI can help you write faster, but it can’t replace the thinking behind good writing

AI Rewards Curiosity

The big takeaway for me this week is that learning AI requires a lot of “what if?”

For example, what if I could build a system that checks specific URLs for new content, summarizes what changed, and gives me prioritized blog post ideas based on my brand guidelines?

What would that look like?

What would I need to learn?

Could I configure it to run every week?

Building a system like this is not easy. Along the way, you might get frustrated with your pals Claude and ChatGPT—just like I did.

Things break. Prompts don’t work. The output is weird. You realize you skipped a step because it was so obvious to you that you forgot to explain it.

This is all part of the learning and discovery process.

And honestly, it’s why I’m having so much fun. It’s been a while since I’ve stretched my brain like this.

You Need to Know the Steps Before AI Can Help

A lot of people start with AI by treating it like a search box.

Ask a question. Get an answer. Decide whether it helps.

That works for simple things. But it does not work as well when you want AI to help with strategy, systems, workflows, or writing that actually sounds like you.

This is something we’ve talked a lot about in the AirOps AI Marketing Cohort. If I want to build an AI-assisted process, I need to know the manual process first.

For example, if I want to create a system to help me update my website, I need to know:

  1. Who the website is for
  2. What I want people to understand
  3. What Good Words Work stands for
  4. What I believe about content strategy, writing with AI, helpful content, SEO, AEO, and AI visibility
  5. What questions people are asking
  6. What topics I do and do not want to write about
  7. What my research, writing, and editing process actually looks like
  8. How much time I realistically have to write, publish, and share content

This is the work that makes AI useful.

Without this context, I can still ask AI to write website copy, suggest blog topics, or build a workflow.

I’ll just get back generic, bland content. AI slop.

The kind of content we’re all reading too much of these days.

And that is definitely not the goal.

Why Content Writers Make Good Content Engineers

This is where experience with the content process becomes really useful.

Writers know that good content does not start with typing.

It starts with questions, audience, message, research, structure, judgment, editing, and knowing what “good” looks and sounds like.

We can also quickly identify poorly written, thin content.

This relies on real-world experience—something AI can’t replace.

The last few sessions of the AirOps cohort have focused on Playbooks and Grids. I’m still learning both, but I can see how powerful they can be. Quill, the onboard AI helper in AirOps, makes it easier to build workflows than I expected.

The same rule applies: you still need to know why you’re creating it, what you want from it, and the manual steps to get there.

Demo Day Is Coming

We have two sessions left in the cohort, and the last one is Demo Day.

I’m really hoping my system gets selected.

I’m a naturally shy person, so demoing something I’ve created is a bit outside my comfort zone.

But if I’ve learned anything these last three weeks, it’s that I can do new things.

Comfort zones need to be constantly redefined.

AI Turning Point

Stay curious. Be ready for things to break. Learn the manual steps. Ask better questions. Then build the system.

AI can help you work more efficiently.

It can help you organize, summarize, structure, compare, and repeat the tasks that take too much time.

But it cannot make up for not knowing your own work.

Do your homework. Get good at your core skill. Then use AI to help you get even better.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

I’m Now a Certified Content Engineer. Here’s What That Actually Means

Week 4—May 27, 2026

TL;DR

  • I’m now a Certified Content Engineer after completing the AirOps AI Intermediate Marketing Cohort
  • The biggest lesson: using AI well is not about finding one perfect prompt
  • Repeatable systems give AI the context, process, examples, and guardrails it needs to produce useful work
  • Content engineering helps make AI-supported writing, editing, and strategy more useful, consistent, and human
  • Human judgment still belongs throughout the entire process

What I learned from the AirOps AI Intermediate Marketing Cohort, why content engineering matters, and why using AI well is about systems—not shortcuts

I’m very proud to share that I’m now a Certified Content Engineer. I earned this certification by completing the AirOps AI Intermediate Marketing Cohort.

It was a packed three weeks of engaging live sessions and many hours of homework in my free time. It was definitely one of those courses where you get out what you put in.

What Is a Content Engineer?

A content engineer builds the behind-the-scenes systems that help good content happen more consistently.

In my work, that means defining the process, prompts, examples, standards, and review steps so AI can support the work without replacing the human thinking behind it.

A content engineer thinks about:

  • What the writer or AI needs to know before creating anything
  • What process the work should follow
  • What “good” looks and sounds like
  • Where human judgment is required
  • How AI can support the process without taking over the thinking
  • How to make the workflow repeatable, teachable, and easier to improve

Content engineering helps make AI-supported content more helpful, engaging, readable, and human. Without it, using AI for writing, editing, or strategy can quickly become time-consuming and frustrating.

A good content engineering process helps AI understand the task, the purpose, the context, and what a strong output should look like.

As a content engineer, I create Skills, workflows, code, and Agents, and I work with tools like AirOps to build the systems behind better content. The goal is to give AI the right context, process, examples, and guardrails so it can support the work without taking it over.

What I Learned from the AirOps AI Intermediate Marketing Cohort

The AirOps AI Intermediate Marketing Cohort helped me understand that using AI well is not about finding one perfect prompt.

It is about learning how to design repeatable content systems: defining the goal, giving AI the right context, building the workflow, testing the output, and keeping human judgment in the loop.

Over the three-week course, I learned:

  • AI works best when it is part of a system, not a shortcut
  • Better inputs create better outputs
  • Documenting your workflow comes before automating it
  • AI needs context, examples, source material, and clear expectations to be useful
  • Human judgment belongs at every stage, not just at the end
  • One good workflow can make the next piece of content, process, or idea stronger

This course changed how I think about AI and how I approach my work. I’m looking more closely at what can be made more efficient, so I can spend more time on the high-value work that depends on my experience and expertise.

I also learned how to use AirOps, which was completely new to me before this course. As part of the cohort, we got an early look at new features before they were released.

What stood out most was the way AirOps helped me see content as a system. 

Grids organize the work. Playbooks define the strategy, rules, steps, and approval points. Quill helps run those Playbooks more like an AI teammate than a blank page, while still keeping human judgment in the process.

A big thank you to Katelyn and Terence for an excellent three weeks. The sessions were well planned, well structured, and genuinely interesting. They were also generous and patient with our questions — including, in my case, a few follow-up emails.

AI Turning Point

I may have written this before, but it’s important enough to repeat.

You can “just use AI.” When you do, you usually get junk. It doesn’t matter whether you’re writing content, code, strategy, or something else. Without context and direction, the output probably won’t be useful.

Everything meaningful you do with AI needs human involvement. There is no one-and-done.

To use AI effectively, you need to put in the time to build the Agents, Skills, workflows, and systems around it.

This is not writing a prompt and pressing Enter. It means documenting your process, determining the right AI path for what you want to do, then building, testing, refining, and reviewing.

AI is only as useful as the context, knowledge, and direction you give it. The stronger the input, the stronger the output. There are no shortcuts.

Curious about how content engineering could help you create better content with AI?

Let’s talk about your process, where AI could actually help, and what needs to stay human.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

When a Cool AI Idea Isn’t Actually Useful

Week 5—June 3, 2026

TL;DR

  • I built an AI-powered Event Tracker to help manage trade shows and events
  • The idea worked, but the workflow created a new problem: people had to use Claude every time they wanted to update it
  • That made me question whether AI was the right solution—or whether a simpler tool, like a Google Sheet, would do the job better
  • The lesson: AI can help build something impressive, but the human still needs to decide whether it’s practical

AI can help you build the thing. But it can’t always tell you whether the thing is worth building.

The Idea Was Good

On Tuesday, I created a Skill to help my colleague manage events and trade shows.

Managing trade shows and events is a massive mental load, with an endless list of decisions and deliverables.


My idea was to create a Skill that would track everything in a checklist. As items are ticked off, a percentage-complete progress bar would move along. Decisions and deliverables are categorized as Event and Human, with everything organized by priority.

The Skill didn’t take long to create and really is quite cool. When anyone in the group types something like, “I’m doing the planning for the blah blah show,” into Claude, the Event Tracker Skill is fired.

Claude goes out and finds the details about the show, updates the tracker, and displays a deliverables checklist that includes contact names, deadlines, hotel reservation information, the nearest airport, booth options, and other key details.



As each item is done and ticked off in the Event Tracker, it’s easy to see what needs to be completed next

Because event planning often starts six to eight weeks out from a show, the Event Tracker needs to save information and reload it when the Skill is used again.

To make this work, the Skill writes to a Google Doc when the Save button is clicked.

Then, when anyone in the group opens Claude and types, “I want to keep working on the blah blah show,” the Skill starts thinking, gets the information from the Google Doc, loads it into the Event Tracker, and away we go.

The final piece is that the Event Tracker drafts a show summary email for approval before sending it to a supplied list of internal show attendees.

Then I Hit the Practicality Problem

Pretty nifty, eh?

Only problem: anyone in the group using the Event Tracker needs to do so in Claude.

So every time you want to update the checklist or see what is outstanding, you need to open Claude, type a prompt like “I want to keep working on the blah blah show,” wait for Claude to do its thing, load the data from the Google Doc, and rinse and repeat.

In the end, not that useful.

Cool. But not useful.

I’m not convinced this is a better solution than a paper to-do list and simply striking things off as they get done.

What do you think?

I’ve got the Event Tracker. I’ve got a Show Summary email. But how do I make these practical tools that don’t force people to use Claude?



When I demoed it to my colleague, the concept made sense.

But this question stood out:

“So I have to run Claude whenever I want to update my trade show checklist?”

I haven’t come up with the answer yet. I’ve got the Event Tracker. This is useful—but what do I do with it next?

I guess I can just ask Claude.

But then again, it was Claude who helped me come up with this idea.

There has to be a better way than a list of to-dos jotted down in a notebook.

Maybe AI Isn’t the Right Solution

Why spend all this time telling you about this?

Because I think it’s a good example of having half an idea and trying to force it into AI.

Maybe AI isn’t the right solution here.

Maybe there’s another existing product we could use. Or maybe it’s just a Google Sheet we create for every trade show and event.

AI, as helpful as it is, is not always the solution.

I think it’s easy to get caught up in forcing AI onto every task, to-do list item, and challenge without fully assessing the root of what needs to be done and the shortest path to the best outcome.

Or maybe this is just what it looks like to be a new content engineer who is still at the intermediate level of using Claude, ChatGPT, and tools like AirOps.

I’m curious to know if you’ve had a similar experience with AI.

AI Resources You Don’t Want To Miss

Here are a few articles I think you’ll benefit from:

I keep signing up for webinars, missing them, promising to watch the recording, and then, well, never getting around to it.

So I’m trying to be more realistic about what I’ll actually attend live and what I’ll need to catch up on after the fact.

Every’s Codex Camp: Our Power User Guide — This one is still on my radar. I’m new to Codex, so I’m hoping it will help me get past some hesitation and better understand how it fits.

Reshift, Refocus & Reframe: The New Search Opportunity with Mike King — This one already happened, but I’m keeping it on my watch list because SEO is not dead. It has changed because of AI. This AirOps webinar digs into what matters most for SEO, content strategy, and measurement.

AI Turning Point

AI wants to be your pal.

If you ask it to do something, it will do its darnedest to do it for you.

Even if there is a better way.

And when you get to the end and it’s not quite right, and you ask, “Why didn’t you tell me this wasn’t really doable?” it will still keep trying to help.

So now, when I give AI my great idea, I try to remember that AI is only the middle of the sandwich. I’m the bread that holds it all together.

I can’t assume AI knows the best approach. It can help me, but I still need to provide the guidance and structure. More importantly, I need to do the research required to understand the best options for getting where I want to go.

I still like the “what if I try this?” approach to discovering what is possible with AI.

But you can’t blindly go all in.

Kind of holds true for all aspects of life.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

I’m Learning AI at Full Speed — and Trying Not to Crash

Week 6—June 10, 2026

TL;DR

  • I asked ChatGPT and Claude where I rank on Every’s eight levels of AI adoption
  • They put me at Level 5: Workflows
  • Cool — but also a reminder that I’m forcing the learning
  • This week, I’m slowing down and focusing on Claude Skills, AirOps, and Claude Code

Turns out learning AI is a lot like racing bikes in Belgium

Stop what you’re doing and copy this prompt into your AI tool of choice.

I found it in Every’s guide to the eight levels of AI adoption, and it’s worth trying:

Based on everything you know about me, including memories, tools and skills installed, and past session history, what level would you say I am at on this guide to AI adoption levels?
https://every.to/guides/the-eight-levels-of-ai-adoption

Where do you rank on the Every scale of eight levels of AI adoption?

According to Claude and ChatGPT, I’m at level 5 – Workflows.

Here’s what ChatGPT said about this level for me:



One caveat: you may be more advanced than many Level 5 users in judgment, but less advanced than some in technical implementation. You are not trying to orchestrate fleets of agents or build engineering-heavy infrastructure. You are using AI as a content systems builder, editor, strategist, and production accelerator. That makes you a strong Level 5 for your field, not a wannabe Level 7.

This made me feel a lot better. Because Level 5 is “primarily the domain of engineers,” and I’m not an engineer or developer.

Level 5 is a good place for me. And as I get more comfortable with Claude Code and Codex, I’ll still be a stronger Level 5. Exactly where I need to be for what I use AI for.

“A higher level isn’t necessarily better.”

Of course, these levels will likely change in six months…


Learning AI at Full Speed

Here’s where I am right now with AI — I’m feeling like there is so much to learn and I’m not getting very far.

In 2007, I moved to Belgium and raced my bike for two months. I raced in a peloton of 90+ women, many of whom were national champions and had been racing their bikes since they were quite young.

I had only been racing for a few years and typically in a field of 15 women. Over those two months, I raced over 18 times. This was a lot of racing! More than I had done in the previous two years.

To say I was in deep is an understatement. Every race something new happened. And as a solo rider without any teammates, I had to learn how to think fast on two wheels.

At times I was extremely overwhelmed. I remember my coach saying “you’re learning so many things right now that your body doesn’t know what to do with them. One day it will feel just like riding a bike.”

This is how I feel about AI right now. I’m trying to do all the things. Create lots of Skills, use Code, figure out Codex, read about Cowork, watch all the AirOps webinars, attend all the Every camps and read all their articles, read everything by Animalz and Superpath, and the list goes on.

It’s a lot. I can’t do it all. I’d like to, but my brain is pedalling frantically and can’t keep up.

So, here’s the plan:

  • Claude Skills – I want to get a really good set created, tested, and updated
  • AirOps – I’m a new AirOps Champion, so I want to get going on a use case I have for this website
  • Claude Code – I started something that, honestly, I can’t remember, so I need to get back to it

I can’t do all of this in one week. I need to give myself some grace and remember that professional development and learning that actually helps takes time.

I’m also setting aside an afternoon to read all the open tabs with articles that I’ve saved. If I’m really good, I’ll take notes at the same time.

This is where I’m at. What about you? What are you planning to learn, read, or play with this week? 

I’m an AirOps Champion

I kind of buried the lede, but I’m very proud to share that I’m an AirOps Champion.

After completing the AirOps Intermediate AI Marketing Cohort last month, I learned about the AirOps Champions program:

  • AirOps Champions are the Content Engineers, strategists, and leaders pushing the AI marketing movement forward.
  • AirOps Champions is the program that recognizes and elevates the operators who are walking the walk: leaning into this new way of working, experimenting, driving real wins for their teams, and advancing their careers in the process.

I wanted to be part of this program because I really do believe in how AI is reshaping marketing, content, and strategy. I like learning about AI and creating processes and systems that help me do more for my clients and myself.

As an AirOps Champion, I can be part of a group that embraces technology and helps people who may be hesitant, understand how AI can support their work while still staying true to themselves.

The AirOps Champions program is not a paid program. We’re all in it because we see the opportunity and want to be part of it. 

AI Turning Point

Eating an elephant is a slow process. Rome wasn’t built in a day. Crawl before you walk. Fall down 10 times get up 11.

Cheesy? Yes — but very true for me and where I am with AI.

I need to slow down and get really good at creating the processes and systems that will help me do my best work. I don’t need to know the ins and outs of Agents, n8n, or Google Apps Script yet.

I need to stay focused on the immediate priorities: Claude Code, improved Skills, and AirOps.

Maybe you’re curious about content engineering or want to know how to create a Skill or Playbook — email me or message me on LinkedIn. I’m always happy to chat and help out. 

Or, if you want to talk about AI and how you can use it in your writing or creative life — I’ve been there, I get it — let’s chat.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

Using AI Well Starts With Knowing the Work

Week 7—June 17, 2026

TL;DR

  • Button-clickers try things, test ideas, and figure out how the work actually gets done
  • Writing is not just writing. It’s research, structure, strategy, editing, testing, and a lot of invisible button-clicking
  • AI can help you write when you know the subject matter well enough to question what it gives you
  • When you don’t know the topic, AI can make weak content sound convincing
  • Used well, AI doesn’t make creative people less creative. It gives them more room to do the work only they can do

AI can help you write, but only if you understand the work first

I tried to write this column last night. I spent two hours on it and got nowhere.

I ended up with a pile of words and ideas that didn’t fit together.

This is frustrating for me. I’m a word person. Writing is what I do.

I like to have this column written and published on Tuesday evening, so I was tempted to keep hammering away.

But at 9 p.m., I shut the lid on my laptop and chilled out with a documentary about ice diving.

I’ve learned that if I force the words, it’s a fast track to dismal sentences and writing in circles.

So here we are, Wednesday morning, a bit behind where I like to be, but taking the advice of a colleague and friend to “give myself some grace.”

We Should All Be Button-Clickers

Button-clickers try things. They experiment. They ask why. They figure out how things work so they can create better content.

They do the stuff that needs to be done so they can understand what people are reading, clicking, ignoring, sharing, and responding to.

All of us should be button-clickers.

I’m a button-clicker, and I think it makes me way better at my job than being a regular old writer.

I learned the term button-clickers from a recent superpath newsletter, where Eric shared some lines from an interview Alex Kracov did with First Round Review:

“One of the biggest mistakes people make with early marketing hires is they have someone who can talk the talk, but when it comes to actually doing things, they have to go hire people to click the buttons.”

Alex continues by talking about how the people you need on your team are the ones who will put up their hand and say they’ll figure it out, learn the thing, and click the buttons “on their laptop that produce marketing.”

I do this every day. I think it’s part of the reason why I’m never bored in my job. I’m always learning something new that can help my clients with their marketing and content strategies.

Because writing is not just writing.

It’s research. Strategy. Structure. Analytics. Web updates. SEO. Testing links. Building Skills. Figuring out how to use tools properly instead of waiting for someone else to hand you a perfect process.

It’s all the non-writing tasks that need to happen to create a content strategy, roll out a website refresh, or build a series of Skills for another team.

A lot of the magic of writing and content happens thanks to the invisible button-clicking.

I think button-clicking and content engineering are closely connected.

Both require curiosity.

Both require you to get your hands dirty.

Both require you to understand the work well enough to build a better way of doing it.

This is why I don’t think AI makes writers less valuable.

I think it makes curious, skeptical, hands-on writers more valuable.

The writers who know how to think, question, research, edit, test, click buttons, and build have an advantage. 

The writers who only use AI to do their work faster are the ones who will get left behind. 

On Not Using AI to Write

When you’re new to a topic or point of view, do not use AI to write your content. 

This is the fast track to subpar, boring, same-same, overwritten content.

AI can do a lot, but when it writes, edits, and researches, it can fall back on familiar AI habits.

It does not favour original thought or outlier points of view. It often chooses the most cited source, even when that source may not be correct.

Sure, you can add guardrails and ask AI to do in-depth research, but this is not the same as reading and understanding the content yourself.

When you don’t know the subject matter well, you don’t know how to look for the holes, the obvious information, the questionable information, or the extra questions you need to ask to provide helpful content.

Weak content slips through because it’s easy to get lulled into believing it’s good.

The sentences are smooth. They roll off the tongue nicely. The content is easy to read.

It’s so easy to be tricked into thinking you’ve got an article worth publishing.

Don’t do it.
Please don’t do it.

There is a time and place to use AI to help you write. 

On Using AI to Write

When you know the subject matter like you know the back of your hand, you can use AI to write. 

But even with this depth of knowledge, you can’t sit back and let AI do all the work.

You must be heavily involved.

You need to provide guardrails, including brand and style guides, sample content, a detailed outline, recommended sources, and explicit instructions on how to write.

You need to rigorously edit and question what it writes.

Push back on claims.

Keep an eye out for AIisms.

And be ready to scrap the piece of content and start from scratch, which isn’t hard when you’re a domain expert.

When you think the piece is ready, read and edit it with a heavy hand.

Remember how easy it is to be tricked by smooth sentences, an extra adjective or four, a string of words that doesn’t really say much, and words on words on words.

One of the biggest tips I can give you when writing and editing with AI is to read your content out loud.

If you stumble over words or notice yourself talking and talking without pausing, these are signs that you have a problem.

Finally, I do one more top-to-bottom read. This helps me spot clunky sentence structure, sneaky missed or dropped words, and awkward flow.

Why do I always drop a to or an it?

You may notice that I use a lot of series of threes in my writing.
This is my style. It is not automatically an AIism, even though some people like to treat it as one.
Not always.
Sometimes this is just how a person writes.

AI Turning Point

A lot of creative people are afraid of losing themselves to AI.

They fear they’ll lose the part of themselves that makes them creative.

Writers, designers, artists, strategists, and other creative people feel the same thing.

This is because what helps us be great at what we do is hard to pinpoint. You can’t easily describe it. It’s something in you.

You feel it when you sit down at the keyboard, pick up a pen, open your notebook, or use your favourite design app.

It’s as if your fingers know what to do.

Words, ideas, shapes, colours, and stories start flowing.

Before you know it, you’ve done it.

It’s hard to explain how it happened, but it happened. And you trust it will happen again.

So what happens if you start using AI in the process?

Will you lose your creativity? Will that well of ideas deep inside you dry out? Will you stop being curious? Will you stop questioning the status quo? Will you stop pushing?

No.

You won’t.

Not if you use AI the right way.

As a writer, I don’t use AI for ideas, brainstorming, or random thinking. I use it for fine-tuning, a critical eye, and the repetitive tasks of writing.

What makes me a great writer is everything I do away from my laptop.

The books I read. The long bike rides. The charcoal drawing. The people I spend time with. The Netflix binges. The podcasts.

All of this feeds the part of me that keeps me thinking, listening, paying attention, and curious.

I’m exposed to so many tremendous thinkers, doers, and believers.

This is what feeds my work.

This is what makes me great at what I do.

And it’s how you can keep hold of yourself while using AI to help you do what you do a bit better, and maybe even more quickly. 

So you can focus on the best part of being creative.

Being yourself in the work.

P.S. Last week I wrote about not trying to learn and do all the things.

I’ve managed to stick to this.

This past week, I’ve been using AirOps to help me give this website a complete refresh. Thanks to Quill in AirOps, I’ve got a doable rollout plan for rewriting my website and adding new pages that fits into the rest of my life.

It’s super refreshing to not feel overwhelmed or like I need to do it all now.

Are you a creative professional who knows AI could help your work, but you’re not sure how or where to get started?

Do you want to get beyond basic prompting, without handing over the thinking that makes your work yours?

Let’s talk.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

AI for Writers FAQ: How to Get Better AI Output Without Sounding Generic

Week 8—June 23, 2026

TL;DR

  • AI can help with your writing, but it should not take over your thinking, judgment, or voice.
  • Use AI where it naturally fits. It does not need to be part of every sentence, every decision, or every stage of the work.
  • Better prompts come from better context. When you’re stuck, ask AI to interview you.
  • Content Engineering helps you build the systems, examples, workflows, and guardrails AI needs to support better work.

A practical FAQ about using AI to support your writing, protect your voice, and build better systems around the work

One of the best things about this weekly column about AI, writing, Content Engineering, and professional development is the questions I receive.

Admittedly, I don’t get a lot of emails or LinkedIn messages — but I get enough to make me sit a bit taller and feel excited about how this column is helping not just me but you as well. 

After all, this is why writers write — to share, help, stimulate, encourage, agitate, support, and be there. 

When writing clicks, it really clicks.

It feels like the writer is speaking directly to you. They understand what you’re thinking about, what you’re trying to sort through, and what you need help making sense of.

That’s what I want this column to do for you.

I want this to be the same for you — but less about the existential questions of life and what it means to be human, and more about how you can use AI in your job as a writer, creator, business owner, or marketer without losing yourself.

This is core to everything I’m doing with AI and sharing with you. Use it, but stay true to what you do.

This column is an example of this. By now, I could hand over some rough ideas to ChatGPT and it could likely write a close facsimile of this column.

But I would feel a bit ick about it.

I did give this to ChatGPT to do a few key things for me: H1, TL;DR, sub-heading, H2s, CTA, and an edit. Otherwise it’s all me. 

So, in keeping with the spirit of asking better questions, I thought I’d turn this week’s column into an FAQ.

Some of these are questions I see coming up in conversations about AI and writing. Some are questions I would ask me. And some are questions I think more writers, marketers, and creatives should be asking.

If I’ve missed something, please message me on LinkedIn or email me.

Think of this as your Dear Abby — but for AI. If you’re scratching your head over the Dear Abby reference, read About Dear Abby.

Frequently Asked Questions About AI, Writing, Content Engineering, and Staying Human

Q. How do I use AI without making my writing sound like everyone else’s?

A. You use AI without making your writing sound like everyone else’s by staying in charge of the thinking. Do not start with “write me a blog post about…” and hope for the best. Start with your point of view. Know who you’re writing for. Know what you want the reader to understand, feel, or do.

AI is very good at giving you smooth sentences. This is also the problem. Smooth does not mean interesting, useful, accurate, or original. If you want the writing to sound like you, give AI real context and be prepared to question, edit, rewrite, and sometimes delete what it gives you.

As I wrote in Using AI Well Starts With Knowing the Work, AI can help you write, but only if you understand the work first.

Q. What should I never hand over to AI as a writer, marketer, or creative?

A. You should be very careful about using AI to write about a topic you do not understand. When you do not know the subject matter well, you cannot easily spot weak claims, missing context, or generic conclusions. This is how thin content gets published. The sentences look fine. The structure looks fine. But the thinking is not there.

You should also never hand over your judgment to AI. Or your taste. Or your lived experience. These are the parts of the work that make your writing yours. AI can suggest, organize, summarize, outline, and even draft.

But it should not decide what you believe or what is good enough to publish.

Q. How much of my writing process should AI be involved in?

A. AI does not need to be involved in every part of the writing process. Use it where it naturally fits. For me, I use it a lot for editing and the fiddly bits of writing: headings, CTAs, TL;DRs, and meta descriptions. These are places where AI can help me move faster without taking over the core thinking.

If it does not feel right, do not use it. Trust your gut. You’re a talented writer who knows how to write a good sentence. AI can help, but it does not need to be part of every sentence, every decision, or every stage of the work.

Q. How do I protect my voice when I’m using AI to edit, outline, or brainstorm?

A. You protect your voice by knowing what your voice sounds like before you ask AI to touch it. Pay attention to your sentence patterns. Notice the phrases you use. Know what you would never say. A brand guide can help, but strong examples usually help more.

You also need to edit with a heavy hand. Read the output out loud. Look for phrases that sound polished but do not sound like you. Watch for tidy little conclusions that flatten the point.

If AI smooths out the parts that make your writing feel human, put them back. Your voice is not something AI protects for you.

Q. How can Content Engineering help me?

A. Content Engineering can help you use AI in a more useful, consistent, and human way. As I explain on my Content Engineering page, AI often gives you generic output because it does not know your voice, your audience, your standards, or what “good” looks like for you.

Content Engineering fixes that by building the context AI needs before you ask it to create anything.

This might look like a brand voice guide. It might be a prompt library, a reusable workflow, an editorial checklist, or a “do not use” language list.

Content Engineering helps you build a system. Not a shortcut. The goal is to use AI to support your process without handing over your voice, judgment, or creative work.

Q. What should I include in a prompt if I want useful, non-generic output?

A. If you want useful, non-generic output, your prompt needs more than a task. It needs context. Tell AI who the content is for, what the goal is, what the reader needs to understand, and what kind of output you want.

You can also tell it what role to play: editor, strategist, skeptical reviewer, or brainstorming partner.

And don’t be shy about asking AI to interview you. Tell it what you want to do, then ask it to interview you about the goal, topic, audience, or angle. This helps you work through your ideas and lets you manage the AI so it knows exactly what it is supposed to do.

The more context you give AI, the better the output.

AI Turning Point

This past week, I’ve been thinking a lot about the difference between using AI to manage the work and using AI to do the work.

I’m still focused on rewriting my website using AirOps. You’ll see I’ve added a new Content Engineering page, and I’m in the process of rewriting the other pages and refreshing my blogs.

This is a big job. And it’s exactly the kind of job where AI can help.

Using Quill, I’ve created two Playbooks. One is helping me build a case study to track improvements in this website’s AI citation rate, visibility, and traffic as I rewrite and refresh the content.

The other is helping me identify which pages to focus on first, which key prompts I want to be visible in, and what I need to do next.

AirOps has my brand kit, knows my voice, understands what my competitors are saying and not saying, and is even monitoring my Google Search Console data.

All of this helps me stay focused on the words. In the background, AirOps is helping with the process, structure, research, and decisions that need to happen around the writing.

I could ask AirOps to write all the content for me.

But a writer’s gotta write…

Have a question I didn’t answer here? Email me. I would love to know what you’re trying to figure out.

And if you need help using AI to improve your content without losing your voice, I can help with that too.

I work with writers, marketers, and business owners on Content Engineering, writing, editing, and building better systems around the work.

And don’t be nervous about signing up for a course that stretches you. You might surprise yourself. I know I did.

How to Use AI in Your Writing Workflow Without Losing Control

Week 9—July 8, 2026

TL;DR

  • AI is most useful when it has clear rules, context, and limits.
  • Without guardrails, it can overwhelm you with information you do not need.
  • Tools like Codex and Cowork can help with writing and content workflows, but they still need human judgment.
  • Knowing when not to use AI is just as important as knowing when to use it.

Codex helped me build a useful content-refresh workflow, but only after I simplified the process and set firm rules

Last week I was researching electrolytes and hydration for our upcoming trip to France. We’ll be in France for two weeks riding our bikes and ultimately taking on L’Etape du Tour de France Femmes avec Zwift on August 6.

This is a 121 km bike ride that ends at the top of Mount Ventoux. It’s going to be a long, hot day out on the bike. I expect it will take me around six to seven hours to complete.

Hydration will be critical, especially if the heat wave doesn’t subside. I have an extra challenge because I don’t have a colon, which means my body does not absorb fluids and electrolytes the same way as you.  

Hence the Googling. I asked something like “How many mg of sodium should I put in my bottle for a long bike ride in France?”. 

The Google AI Overview answer started off really well. But as I gave it more information about the ride and my unique situation, it went off the road.

By the end of the 20-minute-or-so chat, I was more confused than when I started and a bit unsure about my ability to complete the event.

I felt extra nervous when it repeatedly said this about Mount Ventoux, “It is globally renowned as one of the most grueling, iconic, and feared climbs in the history of road cycling and the Tour de France.”

Not helpful Google AI Overview — not helpful at all. 

I share this because it reminds me of the importance of giving AI guardrails and rules. 

When we give it free rein, it tries to help us as much as possible by throwing all sorts of information at us. Some of it we don’t want or need. Much of it clutters the point and can result in frustration, wasted time, and losing focus on the initial goal.

Adventures with Codex

A couple of weeks ago I attended the Every Codex Power User Camp. I had never used Codex before attending this two-hour camp.

But after seeing what the Every team members were doing with it, and frankly feeling incredibly overwhelmed, I wasn’t sure if Codex was for me. 

What helped was when we spent 15 minutes or so in break-out rooms with random Camp attendees.

I was fortunate to be in a break-out room with three other developers. I immediately said I was intimidated by Codex and wasn’t sure it was for me, especially after seeing what the Every folks were using it for.

Without missing a beat, all three told me to just go for it. To open Codex and start chatting with it. So I went for it.

I started in ChatGPT. I told it what I wanted to do and asked if this was a good use case for Codex. Turns out my idea for using Codex to help me with refreshing content is an ideal use case.

I asked ChatGPT to help me with prompting Codex and then I was off to the races. Soon enough, I had a new folder and file system on my laptop, and Codex was working away.

However, I did run into the very problem I experienced when I was chatting with Google AI Overviews. ChatGPT overcomplicated the Codex system. 

So once I had run it from end-to-end, I prompted Codex to tell me how to simplify the process and I told it which files and folders I thought were unnecessary. It took a bit of back-and-forth but now I have a really useful process for refreshing content.

It’s very much a human-in-the-loop process. I point Codex at the original content, tell it what I want changed, and instruct it to create a new draft with labels indicating new and updated content. It alerts me when it’s done. I review and edit the draft, feed my edits back into Codex as context, and then it converts the markdown file to a Word document.

I know I could extend this further by fully automating this process, but I need to be involved and this is not something I need running overnight while I’m sleeping. 

Is Using Codex or Cowork, Cheating?

No. It’s not. Creating these systems to help me work faster and with less friction frees me up to do the heavy, detailed work that comes with writing and content.

Because I’m a subject matter expert in what I’m writing about and only use AI when I know the subject inside and out, I’m confident that the content quality is excellent. Every system I build has writing guidelines, guardrails, checks and balances, and firm rules on needing my input.

I have a client who I’ve been working with for only a few months now. This work largely involves rewriting very thin content. I only use ChatGPT at the very end for editing. I do my research the old-fashioned way with Google. I write the outline and the entire article myself.

I don’t know the subject matter well enough to trust AI with any of it. And because this client is in the YMYL space and requires scholarly article sources and references, I simply don’t trust AI enough for this. 

So while I’m a big advocate of using AI to help with creative and knowledge work, you need to know when to pump the brakes.

It’s pretty easy to identify the content that was written by someone who doesn’t know their subject matter very well and used AI — long sentences that don’t say anything, words that people simply don’t use in normal conversation or writing, jargon on top of more jargon, and an article that leaves you with more questions than answers.

If you’re unsure whether you should use AI to help you research, outline, or write an article, ask yourself this: could I stand up in a room and talk about this subject with a group of subject matter experts? 

If the answer is yes, then use AI. But keep your guardrails in place and stay involved in the entire process. 

If the answer is no, then it’s time to do the writing work yourself. Ultimately, this is what helps you learn the subject so you can become a subject matter expert.

AI Turning Point

Don’t be afraid of it. Codex. Cowork. Claude Code. Vibe Coding. Fable. 

All of it seems very intimidating. 

Especially when you read countless articles and LinkedIn posts from people about how they learned it instantly and, in no time, set up a massive content creation system that is doing the work of 10 people in a fraction of the time.

It’s really hard to wrap your head around this stuff when you’re not technical. I know. I’m in this category.

I’ve found very little practical guidance from Anthropic or OpenAI for people like me. Most of what I can easily find is written for technical or advanced users.

Hey, Anthropic or OpenAI: great opportunity to hire a writer like me to help fill that content gap.

Just try it. Open up Codex or Cowork and type something like, “I’m a [professional writer, marketer, graphic designer, project manager, product manager, editor, or your role]. How can you help me? Interview me about my most common tasks and the way I spend my workday.”

You’ll quickly understand how you can use these tools, when you don’t need to use them, and when a Skill or a simple prompt is enough.

Remember that saying, “If you don’t ask, you don’t get.” Just ask. You never know what you’ll learn.

If you want help using AI without handing over your voice, your judgment, or your standards, get in touch.

I’m a writer and Content Engineer who helps writers, marketers, and teams build better content workflows with AI.

How I Decided Between Codex and Cowork

Week 10—July 14, 2026

TL;DR

  • I’m choosing Codex for my own writing workflow, but Cowork still has an important place in my client work.
  • There isn’t a universal “best” AI platform—choose the one that fits the work.
  • Stop chasing every new release and every AI influencer. Find a few trusted voices and focus on learning.

Choosing the right AI tool—and the right people to learn from and trust—matters.

How to Choose the AI Tool That Fits Your Work

I’ve had one question rattling around in my head all week: Codex or Cowork?



And from what I can gather from chatting with others, Googling, asking ChatGPT and Claude, and reading many articles — I’m not the only one trying to answer this question.



Should I use Codex or Cowork?



Interestingly, when I ask Claude which tool I should use, it tends to point me toward Cowork because it sees Codex as primarily a coding tool.

That is not completely wrong. But it also shows how quickly AI tool advice can flatten the nuance.

Codex is still positioned heavily around coding and technical work, but it is also starting to show up in broader workflow use cases — for example, the Every team uses Codex for their writing and knowledge work.

So, what do you think Codex or Cowork? A mix of both? 



Codex is appealing because I can get it to run at night or when I’m away from my computer. And Codex is a bit stronger in the writing department. 



Cowork is appealing because I have a lot of Skills created already. And I’ve trained it really well on voice, guardrails, and style.

6 of one, half a dozen of another.



While I was out riding on Sunday, I thought a lot about this. I think I’ve come to a decision.

For my writing and website work, I’ve decided to use Codex. I’m a late adopter to Claude for my own personal work, so I have a deep history with ChatGPT.

The history I’ve built up with ChatGPT, combined with the ability to let work run in the background, is what tipped the scales toward Codex.



However, for one of my clients, I’m leaning towards Cowork. The rest of the team is using Claude, we have a a good set of Skills created, and there isn’t anything we really need to run overnight or when we’re away.

Cowork also told me it can continue running while my laptop sleeps. I have a suspicion that’s one of those AI answers that sounds more confident than it should. I’ll find out on Monday.


The good news is that most of the work I’m putting into Codex or Cowork—prompts, guardrails, documentation, and workflows—can be adapted for the other platform. So it’s fairly easy to stay flexible and agile.

There is no right or wrong answer to the question: should I use Codex or Cowork?



AI is moving so quickly that as soon as you settle on one platform, something changes. Last week alone we saw ChatGPT Work and GPT-5.6 Sol announced.

There will always be another release just around the corner.



We also don’t need to chase every new release. The best AI tool is the one that fits the work you’re doing today.



I’m curious — what factors are you weighing in your decision of Codex versus Cowork?

AI Turning Point

I can’t keep up with all the columns, YouTube videos, and posts. I need to choose who I’m listening to and trusting. 



It’s incredibly easy to fall into an AI information vortex that leaves you feeling like you don’t know what you’re doing and are impossibly behind.

I had to remind myself that it’s really easy for people to literally make up stories about how they’re using AI to streamline their workday and their entire life.

They write about this in such a tone that makes me feel like, “dang, how do they do this? They make it sound so easy.” 



Big truth—some people are far more willing to project certainty than admit what they don’t know. And they don’t actually admit the amount of time they put in learning about AI or talking about their failures.

All this to say, evaluate who you’re trusting. If it sounds super easy — they’re not telling you the whole story.

Here is who I’m trusting: 



  • Every – I’m a paid subscriber because they consistently publish thoughtful, in-depth analysis instead of chasing AI headlines
  • AirOps – I trust AirOps because they stay grounded in content strategy and practical workflows, not just shiny new features
  • Andy Crestodina – I’ve followed Andy for years. He brings decades of practical experience in content, SEO, and marketing
  • Tahnee Perry – I’m new to Tahnee, her content is practical and focused on helping marketing teams build repeatable AI workflows instead of chasing shiny new features
  • Nate B. Jones – I’m fairly new to following Nate, but I really like his philosophy of trying things, sharing what he learns, and being prepared to fail
  • Superpath – Is one of the few places where AI is discussed in the context of real content teams and real marketing challenges


If AI has you asking more questions than it answers, you’re not alone. 

If you’d like to talk through content, AI, content engineering, or building practical AI systems, I’m happy to chat.

I’m Making Learning AI Out Loud More Useful

Week 11—August 31, 2026

TL;DR

  • I’m changing the focus of Learning AI Out Loud to make it more practical and useful.
  • I’ll show you how I’m actually using AI, including my Skills, guardrails, workflows, prompts, and outcomes.
  • I want your questions to help shape what I write about next.
  • If you’re struggling with AI, content, prompts, Skills, Agents, Claude, ChatGPT, or simply figuring out where to start, send me your questions.

More practical examples, real workflows, and answers to the AI questions you’re actually asking

It’s been a few weeks since my last column. I was away on holidays and admittedly it’s taken me a bit to get back into the writing groove.

I’ve been thinking a lot about this column and what I want to do with it. 



I really enjoy writing this column. It forces me to think deeply about AI and writing and the relationship between the two. 



But, I’m not sure it’s helping you as much as it could. I think one of the most useful columns I wrote was the FAQ column where I answered the questions you have asked me about AI, content engineering, and content.



So the question is — how to do more of this? 



I think I’ve got a plan. Moving forward I’ll publish a mix of column types: how to’s, lessons learned, and answers to your questions. 



I’ll share Skill instructions, guardrails, workflows, prompts, outcomes, etc. so you can learn from my examples. 



This gives you practical AI guidance that you can apply to your work. 


And I’ll keep sharing what I’ve learned, who I’m reading and trusting, and my successes and failures. 



I’m building my content plan as I type and this is where I want your help.



  • What do you want to learn? 

  • What are you trying to do with AI but not succeeding with?

  • How can I help you? 

  • What questions do you have about AI, content, writing, and building workflows?

  • Do you have questions about writing prompts? Creating Skills? Using Agents?
  • Maybe you want to know how to get started with Claude or ChatGPT?

  • Confused about content engineering and what it is?



There are no wrong questions! I’m here to help you. 


Tell me what you want to learn about.

AI can be very intimidating. Especially when your boss is telling you to use AI and you don’t know how or where to start. 



I felt like this earlier today when I tried to record a Loom. I tried five times and then gave up!



I want you to avoid this and have a positive experience with AI. It’s not going anywhere. This is the time to learn how to use it effectively and efficiently.





Email me your questions. I read and respond to every email. 



Thanks for reading and for being patient with this Learning AI Out Loud refresh.