How to Create Substack Posts with AI

Substack is just as much of a gamechanging platform as AI is a gamechanging tool for any kind of writer. If you’re a journalist with a itch to go rogue and start your own Substack, you can enter the cut throat world of content creation with the cutting-edge in artificial intelligence and gain an invaluable advantage.

I would never tell a professional journalist to use AI tools like OpenAI’s ChatGPT or any other chatbot to generate AI writing for their Substack newsletter. However, there are so many applications for generative AI in a writer's process, that AI can reshape and optimize a writing workflow for life.

Some of Substack's most popular newsletter leverage AI because the most succesful creators are never shy about using the tools provided to them.

Table of Contents

• Using LLMs as Writing Assistants

• Authenticity is Key to Paid Subscribers

• ChatGPT vs Undetectable AI

• SEO with AI

• Conclusion

Using LLMs as Writing Assistants

Journalists and professional writers care about intentionality when trying to engage and connect with their readers. That’s why, if Substack bloggers are uninterested in using AI content in their blogs, they can still utilize a GPT for various other purposes.

AI can take on a role that goes well beyond the capabilities of any writing assistant if a creative is open to it. Using the right AI prompts, writers can be rewarded with solutions like brainstorming ideas for articles, researching data and key information relevant to the topic, comparable articles and authors to learn from, proofreading for spelling and grammar, constructive criticism, and SEO, which we’ll dive deeper into ahead.

Authenticity is Key to Paid Subscribers

Subscriptions are driven to your blog because readers believe in a content creator and want to stay loyal to them. They get a sense of authenticity when reading their words. This is part of the reason why many writers see no point in using ChatGPT to generate text when its writing style feels so inauthentic and robotic.

Pangramlabs collected data about readership of AI blogs vs human blogs and found that readers spend 93% more time on articles that were written by humans and were over 3 times more likely to visit human blogs vs AI blogs.

The graphs above show the distribution of content between human and AI writers as well as which topics are more commonly generated by AI.

Authenticity isn't always about style though, it's about correctness as well. With AI often times generating hallucinations and misinformation in text outputs, it's never been more important to fact-check articles you publish, especially when using AI.

If the originality and authenticity of a piece of text matters to a creator but they feel the need to generate text nonetheless, they should look beyond ChatGPT, Microsoft AI, or Google Gemini, to create content.

Creators need AI text that feels human to engage human readers, and in the generative AI space “human” text is synonymous with “undetectable text”.

ChatGPT vs Undetectable AI

Writers know what makes a piece of text readable is the human qualities of the writing style. ChatGPT has very little humanity in that writing style, automation of writing is meant to generate precise, simple text anyone can understand. Human writing is complex and random and weirdly enough, it’s those qualities that make it feel natural and authentic.

“Undetectable AI” or “Stealth AI” engines use natural language processing to create text that sounds human. It’s the only form of AI content creators should consider usable for Substack, blogging, LinkedIn, social media posts, or any kind of writing meant to be enjoyed.

SEO With AI

High-quality writing with high readability isn’t enough to get people attracted to a Substack page though. Creators need to write with search engines in mind so when someone asks Google a relevant query to their article, it’s their Substack that offers the most useful answers and information.

Google can't be gamed but creators can build articles that are more optimized for people’s searches than others. Writers can do this with painstaking keyword and audience research that can be performed by pricey SEO services, or they can use AI's that are built specifically for SEO.

Beyond finding the right keywords, these AI engines can pillar and cluster the proper backlinks into an article and personalize text to any target audience's voice, a feature which most writers, no matter how talented, don’t have in their bag of tricks.

Being so adaptable that an AI can draft an article written by a young Black girl in one second then an old White man the next is something humans can’t just learn to do. And if a creator needs to write Substack articles for audiences with that wide of a range, using AI is the only way they can optimize as a content creator.

Conclusion

Effectively using AI for journalism and writing is a balancing act between communicating honestly and optimizing the writing into content that is of higher quality than what naturally and immediately comes to a creator. AI collapses time for writers, taking the effort they put into revisions, brainstorming, rewrites, and more and making it take a fraction of the time.

Traditional writers may have an aversion to accepting new technology, such was the case with the advent of typewriters and then computers. However, when the drive and passion for publishing writing that is authentic marries the technology that makes a writer their best self, they will see every metric for success, from subscriptions to views, increase with less effort and more gains.