If you're looking to leverage AI to handle the tedious parts of creation so you can focus on the actual content, this overview is for you. I honestly could not produce the amount of content I do—videos, podcasts, and daily updates—without these prompts. They speed up my entire workflow.
In this post, I'm going to show you exactly how I use ChatGPT. We'll look at everything from the Mac app to Custom GPTs I've trained specifically for my voice. Plus, I'll break down the advanced automation running in my Apple Shortcuts that helps me churn out show notes and summaries in seconds.
This is probably the most important tool in my arsenal. I created a custom GPT called YouTube Sidekick. I trained it on my own videos and voice so it knows exactly how I like to formulate titles, descriptions, and tags. I use this every single time I make a video.
Pro tip: Don't settle for the first title. I usually ask it for "10 more titles like this" to get variations. Since YouTube allows A/B testing now, I'll grab three distinct options to test against each other.
For my daily podcast, I need summaries that sound like a news anchor reading them, not a robot. That's where Summary Scout comes in. It's trained to digest articles and spit out concise, readable paragraphs.
I'm not an expert in APIs or JSON, but ChatGPT is. I have a Shortcuts GPT that helps me build sophisticated automations. I can feed it API documentation—like for the Movie Database or OpenFoodFacts—and it teaches me how to integrate that data into Apple Shortcuts.
For example, I asked it to help me build a shortcut that pulls new movie release dates. It gave me a step-by-step guide on how to use the "Get Contents of URL" action, handle the headers, and parse the dictionary results. I've built dozens of tools this way that I couldn't have managed on my own.
It's also amazing for Regex. If you need to match text—like extracting an MP3 file URL from an RSS feed—just give the feed to the GPT. It will write the exact Regex string you need to paste into the Match Text action.
This is where we combine everything into a mega-workflow. For my show Primary Technology, I have a shortcut called PTS Daily News. It pulls stories from tech news sites, filters them for today's date, and lets me pick the top 5.
Here's the cool part: inside the shortcut, I use the Apple Intelligence "Use Model" action. It sends the article text to ChatGPT with a specific prompt to remove fluff and focus on devices and people. It then appends a clean summary, title, and formatted links directly into a Bear note.
That's a look at how I use AI to keep the content engine running. Whether it's training a custom GPT on your own voice or using it to write code for Shortcuts, these tools save hours of manual work.
If you want to grab any of these shortcuts or try the custom GPTs yourself, you can join the community at shortcut.bot. I stream live builds there and share new shortcuts every week. Give these prompts a shot and see how much time they save you.