SEO Impacts for AI Writing Tools
Discover more from AI Writer World
SEO Impacts for AI Writing Tools
How does Google treat AI-generated or AI-assisted content in its search rankings? Let's find out! Also: latest #aiwriter news, perspectives & research.
This week I’m looking into SEO considerations (search engine optimization) when using AI writing tools. To help with this, I tuned into a webinar by Bernard Huang of Clearscope entitled “How to Rank SEO Content in the Era of Generative AI.”
Now, I’ve always viewed SEO as something of a dark art. I’ve been an online publisher for over twenty years now, but I’ve tried not to obsess over SEO during my career. Mostly that’s because I’ve always written for humans, not machines — and SEO has traditionally been about optimizing for machines (primarily, Google’s algorithms). But it’s also because SEO is a specialized technical discipline and so I prefer to leave it to the experts. With that in mind, let’s see what Huang has to say…
Huang began by referencing Google’s latest SEO update, nicknamed “Helpful Content.” Google says the underlying principle is to “create content for people, not for search engines.” Hmm, well that’s what I’ve always tried to do. But I have a funny feeling that’s not enough, by itself, to rank on the first page of a Google search result. So the question is: what criteria is Google using here?
Google doesn’t tell us the specific criteria it uses to rank content, but it does offer guidance in the form of a list of questions to ask yourself; if you answer “yes” to any of them, you might be penalised in Google’s search rankings. While AI isn’t specifically mentioned, this question clearly references it:
“Are you using extensive automation to produce content on many topics?”
But it turns out, it’s not whether you use “extensive automation,” it’s how you use it that counts.
Earlier this year, Google posted an article outlining its “guidance about AI-generated content.” This article states that Google will reward “high-quality content, however it is produced.” So the company is open to AI-generated content; and even seems to encourage it: “If you see AI as an essential way to help you produce content that is helpful and original, it might be useful to consider.”
Huang’s view on this is that “commoditized AI content” that offers nothing new — i.e. there is no “information gain” — will get penalized by Google. However, if you “direct” AI-generated content in an iterative process with an LLM, you stand a much better chance of adding new information to Google’s knowledge base, and thus ranking in its search engine.
“What we've been recommending to our customers,” said Huang, “is to really think about it as a human-assisted content creation back and forth with the AI model.”
Note: this doesn’t imply that you yourself should add content to an AI-generated article. Indeed, Huang suggests that the entire article can still be AI-generated, as long as you guide it. “So you want to tell the model specifically that certain entities or emerging concepts are things that it should care about and should write,” he said.
Personally, I don’t use AI to generate entire articles — and I probably never will. My modus operandi is to use AI writing tools to help me research and structure, and sometimes make content suggestions. But I think Huang’s advice is also useful for those of us who view AI writing tools more as helpers than creators.
But regardless of who or what writes the article, make sure it adds value and achieves “information gain.” Lastly, here are Huang’s conclusions from the webinar:
Ok, onto the week’s news, opinions and resources in the world of AI writing…
News 📰
Techmeme: Books3, a dataset used to train Meta's Llama, BloombergGPT, and EleutherAI's GPT-J, contains 170K+ books from authors like Stephen King and Junot Díaz (via paywalled article at The Atlantic)
The Hollywood Reporter: AI-created art is not copyrightable, US federal judge says. However, note that AI-assisted works can still be copyrighted. “An application for a work created with the help of AI can support a copyright claim if a human “selected or arranged” it in a “sufficiently creative way that the resulting work constitutes an original work of authorship” […]
Wired: How scientific journals are dealing with AI; “Many journals’ policies make authors responsible for the validity of any information generated by AI.” (via Techmeme)
NiemanLab: A new AP Stylebook entry cautions journalists to avoid “language that attributes human characteristics to these systems.”
Reuters: Advertisers shift to AI; e.g. Nestle: “The engine is answering campaign briefs with great ideas and inspiration that are fully on brand and on strategy […]”
Perspectives 🗣️
VentureBeat: Forget SEO: Why ‘AI Engine Optimization’ may be the future; “If I was a journalist, I would want my articles ingested by all of the LLMs,” says Jeremiah Owyang.
John Maeda, an AI exec at Microsoft, on the need for “friction” in AI design. “Adding friction means making sure that AI users are reminded that they’re not getting perfect answers from a superior alien lifeform.” (via X)
Jon Udell explores whether LLMs should be trusted with marketing copy.
Google co-founder Sergey Brin on why he’s come out of retirement: “This AI revolution is just incredible. I’ve never experienced a time in technology and computer science as revolutionary as this, and I didn’t want to miss out.”
The Guardian: “AI is not a one-time bomb, but a slow burn of devastation that is consuming jobs and culture”; despite the doom-laden headline, it’s a thought-provoking article.
Reference 🧠
Jasper: The Best AI Writers to Try in 2023; needless to say, Jasper is the first tool mentioned, but to be fair there are a bunch of others listed too.
Last Writer: Optimized SEO with AI; good reference if you want to get to the nitty of the gritty in SEO.
See you next week, my fellow augmented writers!