When you take some time analysis what Google AI is capable of, such a generating Wikipedia snippets by summarizing long text sequences (link to paper), you understand that the task of SEO specialists simply armed with guesswork is doomed to fail in an unequal fight.
In this example, we're not only talking about old-school extractive summaries, using sentences from the sources, but abstractive summaries, i.e. brand new sentences which don't copy parts of the inputs.
If you're familiar with NLP algorithms, you can watch this video which "explains" (Not For Dummies though) how Google AI does it.
When you see what Google can achieve in terms of NLP, you can easily imagine how the search team leverages its AI capabilities to determine the relevance of search results and generate the SERPs.
You can discover Google's latest AI developments in Search in this piece published in October 2020.
If you think you can outsmart Google with a few quick & dirty tricks, think twice.
You have to strike the right balance between scalable content production, relevance and vibrancy to get a chance of ranking high in the SERPs. Volume alone won't grant you top positions, even on a high-authority (aged) domain.
Is manual keyword search enough to compete with Google AI?
I'm afraid it can't scale when compared to Google's computing power.
You can use ahrefs or semrush or another keyword tool to draft a shortlist of low-competition keywords and brief your writers to create meaningful posts, of course, but the competitive landscape is always moving in all directions, including at Google's end which constantly adapts the SERPs with its own contributions (featured snippets, image carousels, Google Ads and Shopping section, FAQ accordions, etc.)
In the early days, there were maybe 7 to 10 available organic positions on page 1 for low competition long tail keywords. These days, even in the absence of ads, we're probably talking about 4 to 6 in the best cases.
When assessing the opportunity to develop content on specific topics, SEOs should be armed with the same kind of smart tools as the playground owners in order to maximise their chances of success.
Otherwise it's kind of a fight between cavemen armed with flint spears and unmanned Spaceforce drones.
Are there true AI-powered SEO tools?
"AI SEO tools" only returns 3,880 results on Google at time of writing (AI-powered SEO tools, 2830 results, etc.).
There are some basic AI-components in the off-the-shelf SEO tools ( Frase.io prides itself in enabling "SEO that will still work in 2030" (very nice slogan btw)) but at the end of the day the final creative effort still relies on human writing, which is very difficult to scale in a vibrant way (machine-like human writing gets boring very quickly).
I'm convinced that it would be possible today to generate meaningful vibrant content at scale leveraging AI generative models (GPT-3 & others) WHILE taking into account the almost realtime state of the content market, to evaluate the ranking potential of your content output, comparing apples and apples.
This would require research tools far more advanced than what we have today, no offense to ahrefs, semrush, serpstat, ubersuggest & alikes. Actually we would need a seamless workflow which would blend realtime analysis AND content generation from topic prompts, with the right tone of voice. In other words, true effective scale can't be achieved without analysis & creation blended into a single workflow, if not a tool.
Speaking of scale, improving the accuracy of predictive AI SEO models could actually reduce the flood of automated content since you wouldn't need to produce so many pieces to see what sticks.
AI SEO tools are key to level the playing field in search
In my opinion, AI SEO tools, moderated by human editors, are necessary to create a level-playing field with Google's AI superpowers.
We're only at the dawn of search (Google was founded on Sept 4, 1998, 22 years ago).
Voice search, image search, AR/VR environments, OpenAI / Neuralink and other developments make it even more urgent to equip ourselves with AI-tools to develop our content strategy, from research to production. Surfacing on Google (and other search engines) remains key to any marketing strategy.
There's no point trying to figure out what Big G thinks with antiquated methods. Let's develop the AI tools we need for an equal fight.