Table of Contents
Do not index
Do not index
Oh, hi there 👋!!
Task for the day 📝
🧱 Project: Create a blog post from a voice memo
⏲️ Estimated time: 30 mins
🛠️ Tools: Zapier
Are you ready to control the universe with your voice? 🗣
In the previous lesson we learned how to use ChatGPT to create an SEO strategy and SEO-optimized content. What if you have a great idea for a blog post while you’re away from your computer? Or, what if you hosted a webinar or live event and want to turn that discussion into content?
Our friend AI is here and wants to help. Using AI and Zapier, we can automatically turn half-baked ideas and recording snippets into SEO-optimized content on your blog.
Today you’ll be building a voice-powered content flow.
A voice-powered content flow uses voice memos or video as the start of your content flow. OpenAI’s Whisper transcribes your idea and sends it to ChatGPT for content creation. Zapier pushes that content automatically to your blog for review and publishing. In short, with the workflow you'll be building in this lesson, you'll have a tool to capture ideas in a variety of formats and automatically turn them into content 🦸♀️
In this lesson you’ll learn…
- How to convert ideas into content using OpenAI’s Whisper and ChatGPT
- How to use Zapier to automate idea capture and AI publishing
- How to avoid hitting character limits when using AI
Let’s go 💪!
🔥 Prepare to create
This flow has three steps:
1. Idea transcribed to text
2. Text converted into article
3. Article posted to blog
- Idea to Text
Let’s capture your brilliant ideas and turn them into usable text. I do this often while I am driving my kids to and from school or while I’m sitting in the sauna and thinking up brilliant ideas. 🧖
I can’t run to a computer or even use my phone to write a note, so voice is the solution! A quick voice memo sent to an email address, and I’m done!
Let’s practice!
🎤 Open up your voice memo app and share a half-baked idea you have for a blog post. Think about trying to create an outline in your head: what are the key points or topics that should be covered? What do you want to include or focus on?
Just talk normally. Stream of consciousness is fine!
Now we have to turn this voice into text. That’s where Zapier comes in. Zapier will create an email address we can send the voice memo to, and then send our audio into Whisper, which will transcribe it.
⚡️ Open Zapier and create a new inbound email address. Copy that email address and add it to your contact book as Voice Memo Transcription (so you know what it does).
Once you have an email address, connect Whisper from OpenAI. Whisper will transcribe audio into text.
The File will be the attachment from step one. Then in the prompt section, use the PREP Framework to give guidance on how to transcribe, such as punctuation and line breaks. Not too much detail needed there.
- Text to Article
✍️ Next, we will turn the transcription into a blog post. You will connect ChatGPT on Zapier and follow the prompts. Again, use the PREP Framework to give a prompt that will create blogs that fit the keywords you identified in the last lesson and the tone of voice you want for your content.
Remember there is a limit to how much text ChatGPT will create - known as tokens - so you can also ask it to create an outline or a certain number of paragraphs that you will finish up when you finish the AI Sandwich.
Note that for many blog patforms, you will lose any formatting if the content is not written in HTML. You can ask ChatGPT to create the output in HTML so you keep headlines, bold, etc.
🚨 NOTE: If you have video or podcasts that are already transcribed, you could start your workflow here. Often, I will use Descript to transcribe my content and then create a second version of this workflow on Zapier with a different email address. I then send the transcriptions to that address to skip one step in the flow.
- Article to Blog
Finally, you can push the AI output to a blog draft. Just find your blogging platform and create a post or draft.
You can make the subject of your email from Step 1 as the subject of your blog post. You will want the content of the blog post to be the content from ChatGPT in Step 3.
If you don’t have a blog, or your blog is not connected with Zapier, you can connect Google Docs and edit the draft there.
Remember the AI Sandwich, you will need to edit and update before publishing. Still, not bad to get all of that created for you based off of a single idea from the sauna!
Let’s take a look at this recipe in action...
🔥 Get ready to build!
Let’s get everything in place so you can launch your voice-powered workflow. Follow these steps to get ready:
- Find your Voice Memos (iOS) or Voice Recorder (Android) tool or download one that you like
- Create an OpenAI account
- Create a blogging platform or Google Doc (and have your credentials!) - it will act as the destination for your content
- Create a Zapier account to automate the process
IT’S TIME TO BUILD 🙌!
- Set Up Your Automation on Zapier
On Zapier, create a four step recipe that will:
1. Create an email address that will receive your audio
2. Trigger a transcription with Whisper
3. Send the transcription to ChatGPT to create an article
4. Send the article to your blogging platform as a draft
- Record your ideas
Include key points and details you want in the content
- Send your voice recording
Send your audio to your special Zapier email address
- Edit and Post
Zapier will do the rest of the work for you and spit out your draft blog post. Edit the post, add links and artwork, and publish when ready.
That’s a wrap, we hope you found this lesson helpful!
- If you have any questions, feel free to reach out.
- Want to learn NoCode? Join #100DaysOfNoCode for free.
- Want to learn AI skills? Join #100DaysOfAI for free.
Written by
