Can You Write Blog Posts With AI?

As a busy business owner in need of a blog, outsourcing the writing to AI sounds like a smart solution. Producing high-quality content takes time and money, and AI promises to do the job quickly. However, there are important things to keep in mind when using AI for writing. AI is a tool, not a substitute for human expertise or perspective. And publishing or using text without human involvement has risks. 

If you’re considering using AI to write content, this primer can answer some questions and help you use AI tools to your advantage. 

What is AI writing? 

AI writing refers to using products powered by generative artificial intelligence to write content. These tools rely on large language models (LLMs) that are built with vast amounts of online data to come up with relevant responses to prompts or user queries.

Generative AI services can spit out writing in a number of formats: article headlines, essays, taglines, even short stories. From there, you can tweak the text for things like voice, tone, style, and so on. 

There’s no shortage of AI tools for your business:

• Marketing: AI marketing tools handle everything from content optimization to market research. Some create images from scratch. Others beef up content for SEO rankings. 

• Translating text: AI translation tools help people communicate between languages. A free tool might cover over 100 languages. By contrast, an AI translator might focus more on common global languages and excel at retaining nuance in complex sentences. 

• Business writing: Plenty of AI tools tackle writing for business. AI tools can build pitch decks, while others check grammar and summarize lengthy emails. 

• Creative writing: Though not without its fair share of controversy, AI is increasingly being used for creative writing. A standard LLM can turn out silly limericks. Others aimed at fiction writers, feature a Rewrite button for fixing lines. 

Benefits of using AI-generated content

From offering fresh ideas to summarizing public findings, AI can be helpful for writing, especially at the planning and ideation phase. To get effective results, it helps to be specific about what you want, or get more specific in your requests based on the responses you just received.

• Brainstorming: Give AI a prompt, and it can easily toss out a bunch of ideas to help you get started. 

• Research: For research, AI can function like a search engine, summarizing articles, academic papers, public filings, and other literature on a given subject. 

• Proofreading: There are plenty of proofreading tools, but you can also drop a block of text into a tool and ask it to highlight what to improve. For grammar, there are also AI grammar checkers. 

• Thesaurus: Like a thesaurus, AI tools can offer wording suggestions and drill down by factors like article topic, context, desired meaning, and tone. 

Learn how to use Squarespace’s AI writer

Drawbacks of using AI-generated content

AI can be useful for jumpstarting the writing process. But beware relying on it, especially for factual research and publishing content. Here are some downsides to keep in mind. 

• Risk of inaccuracies: AI tools aren’t always adept at discerning credible from false information scraped from the web. These tools have also been known to make up, or “hallucinate,” information in response to prompts.  

• Plagiarism and copyright concerns: Given that these tools are trained on a vast amount of writing from across the web, there are concerns that it might lift or repurpose some of that work without authors’ permission. 

• Ethical issues and biases: Biases are a serious issue, and these tools exist in a legal gray area when it comes to copyright. Beyond that, there are real concerns, particularly in creative professions, that an overreliance on AI risks undercutting the livelihood of people who make their living by creating content. 

• SEO concerns: By design, AI tools generate an aggregate of information, which means the output can lack uniqueness. Search engine algorithms prioritize experience and trustworthiness. Content that doesn’t share something helpful or new can make it difficult to engage readers or rank in search results.

Start using AI for blogging in 4 steps 

Now that you’ve got a sense of the pluses and minuses of AI writing, the best way to decide whether it’s for you is to try it yourself. Here’s how to get started. 

1. Select an AI writing tool 

Before you start cranking out content, it helps to consider your writing goals. Do you need to craft news updates about your brand? Or publish blogs to optimize your site for SEO? Knowing the answers to these questions will help you choose the right software or decide when and how to use an AI writer for each specific task.  

Here are some other helpful criteria to consider. 

• Content quality: Quality varies from tool to tool, so you’ll want to compare results. Note that you’ll still need to edit the final product no matter how good those results are. 

• Safety and privacy: It’s important to keep your data secure. Business owners should ask about practices like data encryption, security audits, and compliance with industry regulations.  

• Customizability: Some tools use specific prompts for their results, while others focus on getting the job done quickly. You may find that the ability to customize improves content. 

• User-friendliness: The best AI tools are easy to use and to integrate into existing workflows. If the program’s too complicated, chances are you won’t want to use it. 

• Value for money: When evaluating an AI tool, it helps to look beyond the upfront costs. Does it require ongoing expenses for things like training, subscription fees, customization, and integration? Those costs can add up, so be sure to consider whether the cost of ownership is worth the investment. 

2. Enter prompts to generate text 

AI tools require a prompt to generate text, which tends to fall into one of two categories: conversational and structured. Effective prompting can take some practice.

• Conversational prompts: With conversational prompts, you explain what type of content you want and how you want it to sound. There’s no need for specialized prompts or expertise, which makes it intuitive to use. Unfortunately, quality isn’t guaranteed run-to-run, so this isn’t for everyone.

• Structured prompts: Structured prompts refer to giving explicit instructions and examples. The more defined your prompt, the easier it will be for AI tools to focus, which in turn will streamline their results. 

For example, there’ll be a big difference in your results if your prompt is “Write a blog post describing a new release of spring arrangements for a floral design company named Hale,” versus “Write a marketing blog post for new spring arrangements for a floral design company called Hale. Refer to these websites as examples of brand voice, and target an audience of Gen Z and Millennial buyers who value supporting small businesses and sustainable farming practices. Include the search keyword ‘Brooklyn florist.’ Create two alternative versions in different styles.”

3. Generate the content 

Remember, the quality of your prompt will influence the quality of your response. So if you’re unhappy with the results, consider adding more descriptions and direct instructions. 

Here are some other things you can do to improve results.

• Get specific. Asking for a short story is one thing. Asking for a gripping murder mystery in the style of Agatha Christie is another. More specificity means less of a chance of receiving a shoddy response. 

• Describe how to present it. Do you want your results as a summary, a bulleted list, or a film noir-style dialogue? Do you want it to cite sources from the last two years? Saying or pasting an example of what you want to receive can also be helpful. 

• Use “do” and “don’t.” Another way to improve your results is by telling AI what to include and what to leave out. For instance, if you’re writing winter season slogans for a florist that only works with seasonal flowers, be sure to tell the tool to “leave out flowers that bloom in summer.” 

• Ask the tool to clarify. To ensure your AI tool understands your instructions, try asking it to confirm what you’re asking it to do before it starts running. Or ask it to ask three clarifying questions before it starts working on a response.

4. Review all content generated by AI 

Now that you’ve got your content, you’ll need to review it. Ask some questions to help you decide whether it passes muster. 

• Does the text sound too vague or robotic? 

• Is this in my business’ voice? 

• Does this closely resemble copy that’s already out there? 

• Does this use someone’s work without credit? 

• Does this include details unique to my business? 

• Is this something I would have written? 

Plagiarism checkers can help you spot direct copying, but aren’t always accurate. Your best bet is to root it out the old-fashioned way by researching and verifying all sources. 

Here are some red flags to watch for: 

• Repetitive language

• Lack of quotes 

• Lack of citations 

• Lack of anecdotes, opinions, or perspective  

• Outdated and incorrect information

After your content clears any flags, we still recommend editing any blog copy generated by AI writers. The best blog content includes your personal experience and expertise, for the same reasons people are interested in your brand or business—the personality and human touch you bring to it. 

Blog content that sounds like you and reflects your perspective will not only keep visitors reading, but offer new insights. These are both things that search engines factor into rankings, whereas more generic content may not be competitive for search visibility.