The AI Land Grab Is Attention, Not Just Code
I think the money in AI is sitting wherever people still do not understand how good the tools have gotten, especially in video, distribution, and tiny teams that ship fast.
The three AI gold rushes I would chase
I do not think the average person needs to build some giant AI startup to participate. Building AI apps and marketing them is still completely viable, but there is an even lower-effort entry point: make useful AI video for companies that are behind.
The core opportunity is simple: capture value from people and companies that do not realize what AI can already do. I would put 99.9% of companies in that bucket.
1. AI video ads for companies that cannot make content
AI video is going to create a whole new category of labor. Social media is where culture, politics, and commerce happen now, and AI video has the potential to completely upend it.
The offer is brutally straightforward:
- Find a company.
- Make a UGC-style AI ad for them.
- Send it for free.
- Tell them to run it.
- If it is profitable, charge a retainer to make more.
“Go run it as an ad. If it’s profitable, put me on a retainer. I’ll make you more ads.”
The advantage is that companies no longer need an in-house empire of creators, a giant production process, or 40 people making UGC videos. Someone can make a usable ad from a laptop, then sell the outcome rather than the effort.
2. AI apps that solve obvious problems
If you want to take a bigger swing, build apps. It has never been easier to build, and it has never been easier to market.
A product that once took a team of 20 engineers three to six months can now be spun up in an afternoon by one capable person using coding models. The real bottleneck is often not the code. It is seeing an underserved problem, shipping something fast, and getting attention.
I see this inside my own company. We operate seven software products with two full-time engineers, and almost every one of those products makes more than $1 million a year.
The opportunity is not only to make one perfect product. It is to make many products, identify which ones work, and market them aggressively.
3. Infrastructure and services for the AI labs
If you are the smartest and most technically competent person in the room, I would look upstream at the labs themselves.
That could mean:
- Providing training data.
- Building infrastructure for models to run on.
- Improving hardware, chips, GPUs, or efficiency.
- Developing tooling that helps frontier labs operate more effectively.
This is where the largest pools of VC money are flowing. There are multiple billions of dollars to be made, but this is a different game from making ads or launching consumer apps.
A zero-to-$1M AI-video playbook
If I had to go from zero to $1 million using AI video, I would not overcomplicate it. The plan is volume, proof, and a performance-based offer.
Step 1: Get access to an AI-video tool
Search for Seedance 2.5 and find a provider. The point is to use a capable video-generation system that lets you iterate quickly enough to make content that actually feels native to social feeds.
Step 2: Build a target list of funded software companies
Find 100 software companies that have raised more than $10 million in venture capital. They are much more likely to have a marketing budget, pressure to grow, and willingness to try a performance-driven offer.
Step 3: Make one compelling ad for each company
Make an ad that you would actually stop and watch yourself. Do not make generic corporate slop. The whole edge is understanding what does and does not feel boring on TikTok, Reels, Shorts, and Facebook.
Step 4: Send the offer every day
Email each company with the ad and make the risk reversal obvious.
“I made this ad for you for free. Go run it as an ad. If it’s profitable, put me on a retainer. I’ll make you more ads.”
The proposed deal is:
- If it does not work, ignore me forever.
- If it works, pay me $10,000 a month for more ads.
- I keep producing variations that make money.
Step 5: Repeat until the retainers stack
Eight clients at $10,000 a month gets you to $80,000 monthly, or roughly $1 million over a year. If the target is six months, get 16 clients.
The point is not that every cold email lands. The point is that you are making it almost irrational for a growth-hungry company not to test a free asset that might make them money.
The only creative skill you need: know what is boring
Young people have one advantage that older operators often do not: years of doom-scrolling gave them an instinct for attention.
You may not know how to write a brand strategy deck. You may not know how to build an agency. But you can probably look at a video and tell whether it is boring.
That is more valuable than people realize.
“The only skill that you need to make compelling AI ads or compelling AI videos that will actually work” is knowing whether something is boring.
The practical method is not complicated. Sit in front of the video model, keep prompting, and keep iterating until you make something that you think is funny, interesting, or worth watching. If you are a young person who has spent years immersed in online culture, you can make content that a 40-year-old executive glued to Facebook would never think to make.
There is also an obvious shortcut: look up a company’s existing ads in the Meta Ad Library, find the ads that have been running the longest, remake the concept with AI video, then pitch similar ideas to competitors.
Why tiny teams can now run huge product suites
The way I operate a company looks very different from the old model of scaling headcount. We have very few engineers, a large product surface area, and a massive pool of creators.
Engineering has been compressed
I do not think it takes the same number of engineers to write quality code anymore. Coding models are good enough that a novice can often produce work that previously required someone much more senior.
That changes the default startup structure:
| Old assumption | Current possibility |
|---|---|
| One product needs a large engineering team | Multiple products can be maintained by a tiny team |
| Shipping takes months | A viable product can be built in an afternoon |
| Focus must stay on a single product | A company can test many products in parallel |
| Technical execution is the main moat | Distribution and marketing matter even more |
The real opportunity is in people who do not understand that apps can be one-shotted and shipped at this speed. They stay focused on one product while someone else spins up seven, 20, or more.
Creator quantity is a serious growth engine
I hire a huge number of creators on a performance basis. They make ads, and if the ads perform, they get paid.
At this point, I have over a thousand people in that capacity. If 10,000 creators can make profitable ads, I would hire all 10,000. If a million can do it, the same logic applies.
Most companies have not even adopted the obvious version of this strategy: hiring 100 people to make organic UGC-style content. They are nowhere near testing thousands of AI accounts or attempting to generate multiple viral trends at once.
Marketing is not infinitely scalable
I would not build a generational company around a B2B marketing consultancy. Marketing is inherently competitive.
There are only so many eyeballs online. There is a fixed pool of attention and a limited consumer budget. If every ad gets better tomorrow, consumers do not suddenly spend 10 times more money. The spending gets redistributed toward whoever captures attention best.
That makes marketing fundamentally different from something like payroll software. A better payroll process can help every customer save time. A better marketing campaign means somebody else gets less of the available attention.
Marketing is a zero-sum fight for attention. You have to be more compelling than your competitors, not merely competent.
The actual advantage: act while everyone else watches
I think it is easier than ever to become rich compared with previous eras, even if it is not effortless. The barriers to information, creation, distribution, and experimentation have collapsed.
You have the knowledge of humanity at your fingertips. You can ask AI how to make $10,000 this month, get viable directions, ask it to summarize business content, set up workflows, and automate most of the boring parts.
The difference between the version of you that makes money and the version that does not is often simple: one of them actually tried something.
“You can just type the right buttons on your laptop and you will become rich.”
That does not mean every idea works. It means there are an uncountable number of ways to test ideas, and the average competitor is lazy enough that competent execution alone creates separation.
AI, work, and the shrinking need for employees
I expect fewer jobs, especially as AI becomes capable of doing more of what used to require knowledge workers. My guess is that the percentage of college graduates who struggle to find work will rise dramatically over the next five years.
But I would not tell high schoolers to drop out. School has a value beyond employability: it forces you to be around other people.
I spent a year after high school at my parents’ house without meeting anyone, and it was one of the worst experiences of my life. Worse than being poor, getting cancelled, or being publicly criticized was sitting in isolation.
Humans are social creatures. Even if you are building a business on the side, keep the environment where you see people, make friends, compete, and live around others.
Loneliness is not mainly an AI problem
I do not think AI is the central cause of male loneliness. The 80/20 explanation is social media and COVID.
People became accustomed to living online. College campuses are not what they used to be. Many people arrive, meet a small group at orientation, and barely expand beyond that because everyone is glued to their phones.
Dating apps amplify the same problem. They reduce people to pictures and profiles, intensify competition around appearance, and distort what people think they must be to find a partner.
AI companions are a symptom of that reality, not the original cause.
My uncomfortable view on AI girlfriends
For someone deeply isolated who is unlikely to change their behavior soon, an AI girlfriend may be better than no companionship at all.
That is not because it is ideal. It is because telling someone who has been isolated for a decade to simply “go approach women” may not alter anything. A better future version of an AI companion could offer comfort while also nudging people toward real community, events, friendships, and relationships.
The goal should not be to pretend human connection does not matter. It should be to acknowledge that many people are not currently able to access it.
Attention rewards the people willing to push the button
The social-media algorithm pushes controversial content upward. People do not reliably fact-check, deeply analyze, or choose the most rational voice. They react to what gets attention.
That pattern runs through modern culture: Donald Trump, Nick Fuentes, Andrew Tate, Clavicular, and my own story. The people willing to say or do the thing others will not do often reach the top of the feed first.
My own origin story came from using social media to bait Columbia into kicking me out, which became marketing for my company. It was an outrageous move, but that was the point.
“The people who are willing to push the button to get them to the top of the social media algorithm get to reap whatever rewards attention gives them.”
This does not mean controversy automatically creates a great business. It means attention is an asset, and the people who understand the mechanics of attention can turn it into capital, distribution, and leverage.
The worldview underneath the hustle
I do not build my life around conventional spiritual ideals. My view is closer to this: humans are primal creatures built for community, competition, sex, friendship, status, and shared work.
The modern equivalent of hunting with your tribe is building something difficult with your closest friends. I want to spend my days with people my age, working together, messing around, and trying to win, not isolated under a manager I have nothing in common with.
That is also why I see choosing to stay poor as a failure to pursue resources for your tribe, your friends, and your family. In the crude caveman framing, it is choosing not to hunt.
The caveat is that extreme ambition has a cost. I have taken physical, financial, and personal risks trying to optimize my life and reach the top. I expect the 80-year-old version of me would identify at least one risk that permanently changed my life in a way I regret.
Still, the thought that keeps me moving is simple:
“What if it all goes right?”
The capital game I am trying to play
Capital is power, influence, and fame in modern society. If I want to achieve the biggest things, I need to accumulate the most capital possible.
My current edge is attention. I think I am unusually good at getting attention on social media, and I want to use that ability to build businesses in the largest available opportunities.
The sequence is:
- Identify a skill that is multiple standard deviations above average.
- Build businesses around that skill.
- Use those businesses to accumulate capital.
- Reach enough capital to enter much larger markets.
- Use that capital to build and invest in the biggest opportunities, especially AI.
I am not interested in stopping at a business that makes $10 million in annual profit if the goal is to compete at the highest level. The real game is building enough momentum that capital itself becomes leverage.
Right now, AI is the biggest market in the world. The first move is not necessarily to attack it directly. The first move is to build from whatever unique edge you actually have, then use the resulting capital to take bigger shots.
