Feature
The Gist
- Growing use. Generative AI tools are used by 73% of US marketers.
- Human importance. AI can't replace human content creators; oversight and emotional connection remain crucial.
- Balanced approach. Mixing generative AI's efficiency with human expertise is the winning strategy for content creation.
Ever since ChatGPT became a household name late last year, generative AI’s presence in digital marketing has been expanding by the day.
In 2023, most marketers are using some form of generative AI — defined as artificial intelligence capable of generating text, images or other media. A study of US marketers by Statista showed that 73% of respondents reported using generative AI tools, such as chatbots, as a part of their company's work.
Content marketers are clearly interacting with AI via keyboards and screens, but AI can linger darkly in their hearts and minds. It’s only natural to wonder if generative AI tools will become smart enough to make human content creators redundant.
However, despite the existential dilemmas brought on by AI, it’s important to remember that generative AI tools are NOT advanced enough to replace human content creators. AI still needs us as much as we need AI.
“AI is fantastic for pieces of the content creation process,” said Christoph Trappe, director of content strategy at Growgetter and host of the “The Business Storytelling Show” podcast. “But to have AI write articles from scratch without source information from you can be a dangerous game.”
The Risks of Generative AI Tools With No Human Intervention
Generative AI tools such as ChatGPT, Jasper and Copy.ai can put words together competently at lightning speeds based on similar content out there.
But innovative marketers know that publishing content written solely by machines with no human oversight is irresponsible because:
- The copy will be generic, lacking personality and emotional connection to an audience. Also, it’s worth noting that while Google search does not police AI-generated content per se, it does penalize unoriginal, generalized content, a category that AI-generated content falls into.
- Generative AI tools on their own can’t be trusted to get the facts straight. AI has been known to “hallucinate” information that it has not been trained on, resulting in responses that are misleading or completely made up.
“One of the biggest trends we're thankfully seeing is that smart marketers still understand the premium value of human-generated content,” said Robert Rose, speaker, author and CEO of The Content Advisory. “They’re not falling for the idea of outsourcing content to an ‘AI-powered content factory’.”
While it may be tempting to treat AI like a content silver bullet, this approach will only lower the bar for marketing content overall, said Christopher Penn, co-founder and chief data scientist at marketing data analytics company, Trust Insights.
“Subtle wrong details in AI-generated content can tarnish your reputation with experts outside your company,” said Penn. “If machines are doing more and more without human oversight, more omissions will occur and no one will notice that it’s a significant problem — but it will damage your credibility with your audience.”
Related Article: The AI and Content Marketing Paradox: Empowering and Threatening the Future
Implementing Generative AI Tools While Keeping Humans in the Loop
Overreliance on AI-generated content is a risk, but not using it at all is equally risky. You’ll slow down production and burn out human marketers on content grunt work that AI can easily generate. AI tools have their strengths and limitations, and so do humans. To get the best of both worlds, consider the following.
Learning Opportunities
Conference
Jun
10
NICE Interactions Las Vegas 2024
Register
Webinar
Jun
13
How to Unlock Exponential Growth From Your Purchase Funnel
Master the art of leveraging the entire sales funnel by applying the all-powerful Conversion Accelerator Framework.
Register
Conference
Jun
17
Forrester CX Summit North America in Nashville 2024
Register
Webinar
Jun
18
AI in the Ecommerce Marketplace
Uncover how AI transforms Ecommerce, enriches Customer Experience and drives Sales.
Register
Webinar
Jun
20
Beyond the Buzz: Understanding the AI Frontier In the World of CMS
Join the event to dive deep into the practicalities of AI applications within digital and content management ecosystems.
Register
Webinar
Jun
20
Revolutionizing Retail: How to Master Precise, Data-Driven Marketing
Discover how Adobe Workfront can be your catalyst for precision, efficiency and growth
Register
Use AI to Curate Ideas That You Turn Into Great Content
You can feed a generative AI tool with topics and keywords that are most important to your brand and ask it to generate blog titles or other content ideas based on existing content on the web. Within seconds, you’ll have a list of solid content ideas that it would have taken an hour-long “brainstorming” meeting to create.
Next, prompt the AI tool to craft a detailed story outline. After the story is written (by a human!) save time by asking AI to create five social media posts based on the story’s content. The social posts may not be perfect, but human marketers can quickly tweak them into perfection.
Use AI to Find Data Points (but Always Verify Them)
Need data points for your ebook? Generative AI tools will crawl the web in seconds and find statistics related to the topics you’re writing about.
This can be a huge time-saver, but one limitation with having AI tools curate data points is you can’t be absolutely certain that the data points (“37% of consumers say they purchase products via social media”) and the sources of the data (“according to a study by PwC”) are real. Remember what I said about AI tools “hallucinating”?
In my experience, most of the statistics that AI tools return are, in fact, real. But on a few occasions, ChatGPT returned data points that I could not verify after extensive Googling. The rule of thumb is to ALWAYS verify any AI-generated data points with a Google search.
Use AI as a Content Helper and Constantly Refine
Even if you’re not using AI tools to write a fully-formed content piece (and you shouldn’t!), AI will still be generating information for you to use as material — descriptions of companies, summaries of research reports, statistics on industry trends, outlines for blog posts and ebooks.
Whatever text you’re pulling from AI needs to be combed for generative AI’s main shortcomings: factual inaccuracies and bias.
Generative AI’s penchant for biased responses is a technical issue at the algorithm level and will not be solved by marketers. So the onus is on conscientious human marketers to monitor AI-generated content like a hawk to make sure it’s factually correct and not discriminatory against ethnic groups, gender, demographics, etc.
At the same time, humans are on the hook to inject AI-generated copy with personality. People are really good at using wordplay, anecdotes, idioms and humor to connect emotionally with an audience; AI can only mimic this, and badly.
Related Article: Why Marketers Should Stop Worrying and Embrace AI in Content Marketing
Winning Combo: Generative AI Tools and Human Expertise
Any content marketer would be crazy not to capitalize on generative AI’s ability to scour the web for information in seconds, come up with content ideas and streamline repetitive tasks.
But unless your goal is to publish unoriginal, inaccurate and possibly biased content, marketers should not hand over the content creation keys to AI. The winning combo for the foreseeable future is to let AI tools handle data analysis and low-level content generation, and let emotionally intelligent humans produce and refine the final content.
“Generative AI is exceptionally good at writing confident content, even when it is confidently incorrect,” said Christopher Penn. “So the most important best practice for human oversight with AI is to always have subject matter experts reviewing AI output.”
