Using AI Tools to Generate Content? Here’s How To Edit the Output So It’s Usable

Using AI Tools to Generate Content? Here’s How To Edit the Output So It’s Usable

With all the recent excitement around ChatGPT and other generative AI tools, you might be thrilled about the possibility of using artificial intelligence to help you create marketing content.

But how much do you need to edit the text you get from AI content tools, before you can publish that output online?

The short answer is: You’ll need to do significant editing. The raw output from AI software is not publication-ready.

Generative AI can help you create simple first drafts for certain kinds of copy (like social media posts, emails and press releases) but if you publish the content straight from AI software, you’re going to end up neck-deep in problems.

Read on to find out what human editors need to consider when reviewing text from AI content tools, and the biggest risk of published AI-generated output that hasn’t been properly edited.

Here’s What’s Missing From AI-Generated Output

Generative content can give startups additional power to scale content creation — but when you receive the output from tools like ChatGPT, it’s going to be missing some essential elements.

A skilled copywriter or copy editor will need to take the raw AI-generated content and add in:

  • Nuance: Output from AI content tools won’t have an understanding of the nuances of your audience, your company or your products. No artificial intelligence will be able to replicate your team’s experience and expertise.
  • Narrative: Human beings love stories. The lizard part of our brain recalls ancient days when we all sat around campfires telling stories. As of right now, AI can mimic storytelling structure — you’ve probably seen some examples of this online — but machines can’t yet replicate the natural storytelling and narrative skills humans bring to the table.
  • Soul: AI-generated content is based on a massive amount of data. That means it's inherently generic, because it is limited by the data it’s been trained on. Human beings need to add soul and personality to content to really make it come alive.

Your content team will also need to edit the draft for voice, structure and flow. Does the content make sense? Does it provide a positive user experience? Make sure you’re examining every piece from the reader’s point of view.

A Giant Word of Caution About Fact-Checking

You absolutely, positively must fact-check every piece of raw output you get from AI content tools.

The bottom line is that generative AI can lie, and do so in an authoritative tone that sounds incredibly convincing. Tools like ChatGPT will often just make something up, sometimes backed up by fake citations.

Ethan Mollick, a professor at the University of Pennsylvania's Wharton School of Business, told NPR, "The best way to think about this is you are chatting with an omniscient, eager-to-please intern who sometimes lies to you."

So plan on using a fine-tooth comb to fact-check every word of the output you get from the AI content tools you use. That includes any companies, products, websites, links and sources.

Fix all inaccuracies, remove errors and cite proper sources before moving on to the next phase of your content production process.

The Biggest Risk of Publishing Unedited Content From AI Tools

There are a lot of marketers at startup companies who won’t take the time to properly edit the output from AI chatbots and other content-generating tools. And that’s a huge risk.

Why?

Because content that is inaccurate or lacks nuance, narrative and soul will lead straight to broken trust with your audience.

And once your prospects and customers lose trust in your content, they lose faith in your company — and you won’t be able to get that trust back.

With Great Power …

I’m a fan of using generative AI to help with content creation, and I think this new technology can be an especially powerful tool for startups.

But with great power … comes great responsibility.

To maintain trust with your audience and avoid putting customer relationships at risk, you must plan to edit and fact-check any text you get from AI content tools. Build this editing and fact-checking into your content creation process, and make sure your copywriters and editors have the tools they need to work with AI successfully.


For more information on how startup marketing teams can leverage AI for content generation, check out AI for Tech Startups: A Guide to Using AI to Write

This is an insightful and informative post about the use of AI content generation tools. You've accurately highlighted the potential benefits and limitations of these tools. I particularly appreciate your emphasis on the importance of human editing and fact-checking. It's crucial to remember that AI is not a substitute for human creativity and critical thinking. By combining the strengths of AI and human intelligence, we can create high-quality content that resonates with our audience.

回复
Todd Jones

Storyteller | Brand Whisperer | The About Page Guy ? |"I help brands uncover and articulate the stories that make them memorable."

2 年

Nuance, Narrative, and Soul. Agreed! AI misses these. I would also add that I think it may be hard for AI to use subject matter experts. At the least, you will need to examples and SMEs as well as your own opinions. Great post Jessica Mehring!

回复
Sean Work

Digital Marketing Specialist: Expert In Search Engine Optimization & Written Website Content.

2 年

Yep! I tried ChatGPT out for some biographical content for one of my sites and ChatGPT started to vomit out dupe content like crazy. Basically writing the exact SAME content verbatim for each person. I'm not even certain of it's accuracy in terms of sourcing facts. It's possible that this tech will make content creation even more time consuming or end up causing more problems for people who blindly use it.

回复

要查看或添加评论,请登录

Jessica Mehring, M.A.的更多文章

社区洞察

其他会员也浏览了