[00:02]
Google [music] Anti-gravity to build a whole blog that ranks on Google and gets quoted [music] by ChatGpt. Not just one article, the keywords, the content, the design, and all the technical stuff underneath, including the schema that
[00:16]
tells AI exactly what your page is about. I gave anti-gravity one message. It went into my real search [music] console data, found the questions people are already searching, planned the whole site, and then built it. homepage, topic
[00:30]
hubs, articles with direct answers at the top, schema on every single page, a site map, an AI crawler file, and a page that checks its own work. In this video, I'm going to show you exactly what it built, the exact message I sent, and how
[00:45]
to do the same thing for your own business. And stay with me because near the end, I take the same system and point it at a completely different business, a real estate team, in one go. By the end of this video, you'll know
[00:57]
how to turn the questions your customers already ask into a blog that AI search wants to quote. Let's get into it. To make all of this easy to follow, I built it as a place. It's called the glass press. Picture a newsroom made of glass
[01:13]
floating above the clouds. Down at the bottom, minor robots dig glowing crystals. And every crystal is a real search someone typed into Google. In the middle, there's a giant printing press. And running it is a small paper robot
[01:27]
called Folio. Folio takes the crystals, feeds them into the press, and finished articles roll out the other side. Then the AI search engines fly in. Chat GPT, Gemini, Perplexity, that's the whole game. Now your page has to be the one
[01:43]
they pick. Let me start with what actually happened. On the 18th of November, Google released Gemini 4 Neon inside Google Anti-gravity. Anti-gravity is Google's app where AI agents build things for you. Neon is the new middle
[01:59]
model between Google's fast hydrogen and its deep carbon. Google says it ranks number one on the artificial analysis coding agent index and number one on Zapia's automation bench, which tests how well an AI handles real automation
[02:14]
work. Google also says it has the lowest rate of making things up of any model they've tested inside anti-gravity. That matters a lot for SEO because madeup facts lose trust fast. And there's one more thing that came out on the same
[02:29]
day. It's called Orchestrator. Orchestrator lets Neon be the planner while Hydrogen does the fast building work underneath it. In plain English, Neon is the editor and Hydrogen is the team of writers. And if Hydrogen gets
[02:43]
stuck, it hands the problem back up to Neon. Neon fixes the plan and the work carries on. So, every build in this video ran on Orchestrator with Neon as the planner. It's rolling out today on Google AI Pro and Ultra Plans in the
[02:58]
desktop app, the command line, and the editor extensions. Now, before I show you the builds, if you want this whole system for your own business, this is exactly what we do inside the AI Profit Boardroom. You don't just get the
[03:11]
prompts. You get the complete AI search blog system from this video. The keyword engine that turns your search console data into a plan. The blog builder that designs the site, writes the answer first articles and adds the schema. The
[03:25]
schema X-ray, the content rules, and the technical checklist. The skills and agents that run it inside anti-gravity and the agent OS so your AI tools work from one dashboard. the full automation template so you can run it again for
[03:39]
every new topic or every client, step-by-step training videos for every part, and a 30-day road map that takes you from your first export to a published blog that's built to get quoted by AI. So, instead of guessing
[03:52]
what to write, you start from searches you're already close on. Instead of a day per article, you get a whole site in one session. And instead of hoping AI search can read your pages, the schema and the crawler rules are done for you.
[04:05]
Inside the boardroom, you also get weekly coaching calls where you can share your screen and get help on your own site, daily new tutorials as these tools change, and a community you can ask anything, anytime. And if you want
[04:17]
to learn how to make videos like this one, the whole faceless video system is one, the whole faceless video system is in there, too. More than 3,700 business owners are inside right now using AI to save time and get more customers. The
[04:30]
link is in the comments and description or go to aiprofitboardroom.com. And if you'd rather have it done for you every day, try Polarank. Polarank is a team of AI agents that writes your articles, builds your links, and checks
[04:43]
whether Google, Chat, GPT, Perplexity, and Gemini are naming you. It publishes straight to WordPress, Shopify, or Web Flow. and you can try it for 3 days at That link is in the description and comments, too. Okay, let's go down to
[04:59]
the bottom of the tower where the miners dig. Step one is the keywords. Most people start a blog by guessing. They sit down, think of a topic that sounds good, and write about it. Then, nobody searches for it, and the article sits
[05:13]
there with zero visits. Here's the better way. Google already tells you what people are searching for when they find you. It's in Google Search Console and it's free. Search Console shows you every search where your site appeared,
[05:26]
how many times it showed up, and where it ranked. The gold is in the searches where you show up on page one or two, but not at the top. Positions 4 to 20. You just haven't given it the page that wins. So, I took the real search console
[05:43]
data for my own site, aprofitboardroom.com. You can see it here inside my agent OS on the SEO tab. I pulled out every search sitting between position 4 and position 20 and I handed that list to
[05:56]
anti-gravity with one job. Clean it up, group it into topics, turn every keyword into the real question a person would ask chat GPT and tell me which ones to build first. Here's what came back. It merged the spelling variations together.
[06:11]
So, two ways of writing the same tool name became one topic. It grouped everything into clusters with one main page for each cluster and supporting articles underneath it. It turned short keywords into real questions. A keyword
[06:25]
like is Jev open source became a clear question that an article can answer in the first two sentences. And it scored every single one. Quick wins, things to build next, and long-term plays. Then it built this a keyword constellation.
[06:41]
Every glowing crystal is a topic cluster. The smaller crystals orbiting it are the questions. The size is how many people search for it. The color tells you what to do first. Amber means quick win. Blue means build it next.
[06:55]
Click anyone and it flies you straight to it and shows you the numbers. One of the things it flagged was a search for AI SEO agency. That's showing up thousands of times and we're sitting around page two. That's exactly the kind
[07:09]
of page that's worth building because the demand is already there. This is the part I want you to copy. Don't start with what you want to write. Start with what people are already searching, where you're already close. Now, let's go up
[07:23]
to the printing press where Folio works. Step two is the blog itself. I gave anti-gravity one message, and I'll put the full message in the boardroom so you can paste it. It said, "Use my real search console data. Pick the three
[07:37]
strongest topics and the six best questions. Then plan the work and build it all. The design, the pages, the content, and all the technical search stuff. This is where Orchestrator really earns its keep. Neon read the data and
[07:52]
made the plan first. Then it split the build into parts. So, the design, the pages, the articles, and the technical files could all be worked on. And it checked how they fit together at the end. And here's the blog. had built a
[08:06]
homepage with a big hero, the latest articles and the topic hubs, three hub pages, one for each topic, six full articles underneath them, an author page, an about page, and every article links to its hub, and every hub links
[08:22]
back to its articles. That linking is a big deal. It tells Google and AI search how your site fits together and which page is the main one for each topic. Look at the design for a second. big headlines, warm colors, clean cards, and
[08:37]
a proper dark mode. It looks like a real magazine, not a template. And it's fast because it's plain pages with no heavy add-ons slowing it down. Step three is the content. And this is where most AI blogs fall apart. Most AI written
[08:53]
articles are long, vague, and they bury the answer at the bottom. AI search hates that. When someone asks ChatGpt a question, it's looking for the clearest, most direct answer it can quote. So, every article on this blog follows the
[09:08]
same rules. At the very top, there's a short answer box, 40 to 60 words. It answers the question straight away in plain English. That's the bit AI search loves to lift. Then, a list of key takeaways. Then, headings written as
[09:24]
questions, the same way people actually ask them. a comparison table where it makes sense, step-by-step instructions, a question and answer section at the bottom with five common questions, a date showing when it was last updated,
[09:37]
and an author box so it's clear a real person stands behind it. Now, here's the part I really like. Anytime a fact needed a source that the AI didn't have, it didn't make one up. It left a clear note saying, "Add your source right
[09:52]
there in the article." That's the low fabrication thing I mentioned earlier, showing up in real life. You fill in those notes with your own data and the article becomes stronger than anything a competitor can copy. Step four is the
[10:06]
technical stuff. And I know this is the part people skip because it sounds scary, but it's actually the part that gives you the biggest edge because almost nobody does it properly. Let's start with schema. Schema is a small
[10:20]
block of code on your page that describes the page to search engines in their own language. Your readers never see it, but Google and AI search read it to understand exactly what the page is. This is an article. This is the author.
[10:35]
This is the company behind it. These are the questions and answers. These are the steps. Anti-gravity added schema to every page on the site. It describes the organization, the website, the author, each article, the breadcrumb trail
[10:50]
showing where the page sits, the question and answer section, and the step-by-step guides. And it's all joined together in one connected map so search engines can see how every piece relates. Then it added the rest of the technical
[11:04]
list. A site map, which is a list of every page, so search engines don't miss any. A robots file, which is the welcome mat at your front door. This one specifically lets in the AI crawlers, the ones from OpenAI, Google,
[11:18]
Perplexity, and Anthropic, so they're allowed to read and quote your pages. An LLM's file, which is a short, plain English summary of your whole site written for AI. A proper page title and description on every page. Sharing
[11:32]
previews so your links look good when someone posts them. And a custom page for broken links. Then it built one more thing I didn't expect to love this much, a report page. It opens every page on the blog, counts the words, finds the
[11:47]
schema on each one, and counts the internal links so you can check the whole site's health in one place instead of opening pages one by one. And because schema is invisible, I wanted you to actually see it. So I asked Antiggravity
[12:02]
for one more build, a schema x-ray. It reads the real schema from the blog and turns it into a 3D map you can fly through. The organization sits in the middle. The website connects to it. The author connects to the articles. The
[12:17]
articles connect to their questions and their steps. This is exactly what Google and Chat GPT see when they look at your page. Not the colors and the fonts, this map. If this map is messy or half of it is missing, AI search has a much harder
[12:32]
time trusting you. If it's clean and connected like this, you're making their job easy. Now, let me show you the part that makes this a system and not just a one-off. I took the exact same approach and pointed it at a completely different
[12:46]
business, a small real estate team that helps firsttime buyers and sellers. One message, plan the keywords first using the real questions buyers ask chat GPT. questions like how long buying a home takes and whether you should sell before
[13:02]
you buy. Then build the blog, the articles, the team page, the schema, the site map, the robots file, and the LLM's file. And here's what came back. A completely different design. Warm cream, deep green, gold. It feels like a calm,
[13:18]
trustworthy property firm, which is exactly right for that business. four articles, each with the answer box at the top, the steps, and the question section. And the schema is set up for an estate agent this time, not a software
[13:32]
blog, because Neon understood what kind of business it was. That's the bit that saves you real time. You don't rebuild the system for every client or every business. You change the business and the system adapts. So, let's put the old
[13:46]
way and the new way side by side. The old way, you guess at a topic. You spend a whole day writing one article. You hire a designer to make the blog look good or you live with an ugly template. You never get round to schema because
[14:01]
nobody explained it. You don't know if AI crawlers can even read your site. And you wait months to find out whether anyone cares. The new way. You start from the searches you already show up for. Neon plans the whole site from that
[14:16]
data. The design, the hubs, the articles, and the technical files get built in one session. Every article opens with the answer AI search wants to quote. Every page has connected schema. The AI crawlers are invited in on
[14:31]
purpose, and there's a report page that checks it all for you. Now, you might be sitting there with a few doubts. The first one is, isn't AI content bad for SEO? Google has said it cares about whether content is helpful, not whether
[14:46]
a human typed it. The problem with most AI content is that it's vague and made up. This system does the opposite. It starts from real searches, answers the question in the first sentence, and leaves a note instead of inventing a
[15:00]
fact. Then you add your own experience and your own data on top. That combination is very hard to beat. The second doubt is, I'm not technical. I'll never get the schema right. You don't have to write a single line of it
[15:14]
yourself. That's the point. You ask for it in plain English and the report page shows you it's there on every page. The third doubt is does this only work for software or tech blogs. You just saw it build for a real estate team with a
[15:29]
completely different style and a completely different type of schema. If people in your industry ask questions before they buy, and they do, this works for you. So, here's what you can do this week. Step one, open Google Search
[15:43]
Console for your site. Go to the search results report and export your searches. Filter for anything sitting between position 4 and position 20. Step two, give that list to anti-gravity with neon orchestrator with neon as the planner
[15:58]
and ask it to group the searches into topics and turn them into real questions. Step three, ask it to build one topic hub and three articles first with an answer box at the top, a question section, and connected schema
[16:13]
on every page. Don't try to build 50 pages on day one. Step four, add your own experience, your own examples, and your own sources to those articles. Then publish them and check them in search console a few weeks later, one topic at
[16:28]
a time. And if you want help doing that, join the AI profit boardroom. You get the whole AI search blog system from this video ready to use. The keyword engine, the blog builder, the schema X-ray, the content rules, and the
[16:42]
technical checklist, the skills and agents that run it in anti-gravity, and the agent OS, the automation template so you can run it for your next topic, your next site, or your next client, the training videos, and the 30-day road map
[16:56]
that walks you through it one step at a time. That means fewer hours writing, a blog that's built for how people search now, and a system you own and can reuse. You also get weekly coaching calls, daily tutorials, and more than 3,700
[17:11]
business owners you can learn from. And if you want to learn how to make videos like this one with the 3D worlds and the motion design, that whole faceless video system is inside, too. If you'd rather not do any of it yourself, Polar Ranks
[17:24]
agents will write, link, and track for you every day at polarank.ai. The link is in the comments and description, or go to a profitboardroom.com. So, that's Gemini 4, Neon and Google
[17:38]
Anti-gravity for AI search. Neon is Google's new middle model, built for real work with a very low rate of making things up. Orchestrator lets Neon plan while Hydrogen builds. Start your keywords from search console, not from
[17:53]
guesses. Put the answer at the top of every article. Add connected schema to every page and invite the AI crawlers in. And once it works for one business, point it at the next one. Drop a comment and tell me, would you start with the
[18:06]
keywords you're already close on or with a brand new topic nobody in your niche has covered yet? I'll see you in the next one.