The peak article I linked above is not the usual generic stuff, it's super super detailed and accurate.
So, give it a read to test something new about internal linking.
I know, some of this stuff is too technical and sounds like technical jargon.
I will give you some examples in my future threads...
Anyway, I've seen the results of extreme Semantic SEO several times, given you have the budget.
I want to make one more point clear since many of you will eventually ask this.
Google doesn't expect perfection, that's why you can excel in some scenarios without that much knowledge or just by writing "good content".
In some countries you have search monopolies, meaning that there has to be some topical relevance involved.
If in the US you may complain about those 3-4 large publishers, in European countries you can notice how some publishing firms have the de-facto control of markets.
You can outrank them if you have a more relevant offer but...
Their amount of content is too large that they will have some relevance.
Some even claim that backlinks can fix a lack of relevance. True in some cases and good luck with future updates lol
You see hundreds of cases but no one can prove they're a representative sample
I guess that the main reason behind the success of some publishers is that they just do it.
This will probably change soon, as there is more awareness and defending a niche is quite hard.
Semantic SEO offers a way to save SERPs from harmful results, but we're not there yet.
AI can be dangerous if biased. And let's be fair, modern AI is.
Google's technological advancements are impressive, barring a few niches where there are anomalies.
Generating text will probably be the next step for Google, as heavily implied by their doings.
As of now, AI's just a baby and the first thing SEOs think of are those overpriced tools for generating text.
There are many use cases actually, as I said before, studying some patents can help you in getting those clues.