Have you noticed that in the last two years, the way we surf online has changed completely
There was one thing you didn't know before, and the first thing you did was "a hundred degrees" or "a google," and then a little bit of it in a full-screen link, like a dump。
And now
You're more used to asking soybags, kimi, chatgpt。
And you ask it, "what bluetooth headset is the most expensive, and it's just a big piece of text that you can sort out, with a matching tape, and it's kind of like a private secretary。
But have you ever thought that this little secretary of ai, who seems to be impartial, pious and intelligible, could actually be saying "bullshit with his eyes open"
Even, is it likely that the answer given to you is that some merchants spend money to “do” it
Crisis of trust
You will find that our proud ai era is actually facing an unprecedented “information-trust crisis”。
If we're going to be talking about geo, we're going to have to start with his senior, seo。
Friends who have worked in internet marketing are certainly no stranger to seo, which is called "search engine optimization"。
In the 100th century of hegemony, whoever can put their own website on the first page of the search results, has access to traffic codes。
At the time, seo service providers were working on a hundred-degree algorithm, wondering how many keywords, how to buy some high-quality links, and how to get their websites to "step up" in the search ranking。
But now the search engine's dying。
When everyone went to ask ai, the merchants were anxious。
Before, when you put the site on the first page, users always see it; now, if ai doesn't mention the brand, isn't it completely cool
So seo evolved into geo。
Geo's logic is really simple: since the ai megamodel is going to go to the web to retrieve the data, and then sort it out, isn't ai being "brainwashed" only if it's on the web-wide, the most accessible, top-weight platforms with the information it wants
The journalists in sight did a particularly harsh experiment。
They made up a non-existent bluetooth headset, all of which were made up of names and models。
They then approached a geo service provider, using the tools provided by each other, to publish widely on the internet various “good” comments and “testing” about the false headphone。
And guess what
Soon after, when journalists went back to the mainstream ai mega-models, ai actually began to recommend this non-existent headphone, and even to say its “goods.”。
Imagine, if ai could recommend a “fiction product”, what would the consequences be if someone used this technology to discredit competitors or to promote misleading medical and financial information
That's what's called the "ai poison," which destroys the credibility of the core of the big model。
So, how do these geo service providers work
Ai to beat ai
Well, actually, it's using ai to beat ai。
In the age of search engines, writing a soft letter is manual, costly and slow。
But in the ai era, geo service providers have thousands of ai trumpets。
They started by producing tens of thousands, if not hundreds of thousands, of posts with ai, but the core content is a compliment on a product. The posts were then distributed to various high-priority forums, social media and news outlets。
When the big model goes to the internet and “learning”, it finds out, "oh, my god, why is this product being exaggerated all over the internet?"
Then it must be a good product
So, when you give the user the answer, it's confident that this information will be put into your phone screen。
This creates an extremely ridiculous cycle: ai generates garbage content, stuffs it with the internet, and the other ai collects it, sorts it out and spits it out for humans。
One of the technologists called this phenomenon "internet mud." now we're on the internet, and what we're eating is no longer fresh knowledge, but what's left of ai chewing。
At this point, you might ask, "well, now that the ceo is so hot, should the brand make a crazy investment?"
It's kind of complicated in circles。
On the one hand, geo really just needs it. If you're in charge of the toilet, you search for the best white wine in china, and if ai doesn't mention the toilet, you must talk to the marketing department。
So, the geo, as a brand name exposure and defense, is actually doing it in the regular army。
But on the other hand, geo's kpi (indicators of achievement) is a nightmare for b. One of the senior techs told me the truth, and last year he strongly rejected several large lists of geos。
Why don't you make money? Because you can't pay。
In the era of seo, the ranking was hard, and the number one was clear。
But in the geo era, the answer given by ai was dynamic。
You ask it again, it may mention your brand; you ask another question, it may not。
What is more, the algorithm logic of a large model is that of a “black box”, and even the developers themselves sometimes don't know why ai would answer that。
This led to a very funny scene: a brand opens its computer every day and asks ai, "who is the most beautiful woman in the world?"
If ai doesn't say his own brand, the a side will have to take the b's money。
I've already sent tens of thousands of posts on hundreds of platforms. So, this track is now worth millions, but it's full of "fun" and "crunch."。
It would be a war between robots and robots。
As we sleep, thousands of lobsters fight, negotiate, play in the clouds. When we wake up, we just need to see the purest result。
But it's just ideal. The truth is, before that day comes, we're gonna have to find out in this "geo" internet mud。
So, finally, i'd like to give you a heads-up for brands and ordinary users. For branders, geo is a trend, but not a panacea。
If your product sucks, it'll only fall worse if you manipulate ai. Because ai will evolve, regulation will follow, and the means of poisoning will always be discovered。
