The Only 9 Ways To Make Money With AI in 2026 transcript

Every AI money video makes it sound like there is a secret button to print money with AI. I have used AI inside my businesses for years. I've used it to increase our company's revenue by 50,000 a month. I've used it in ways that have lost us thousands a month. [music] So, in this video, I am ranking the only ways I would actually consider making money with AI in 2026. Not the easiest sounding ones, not the ones gurus are selling us, the ones that I've actually done myself [music] and that connect to real obvious demand.

I'm going to rank each one on a tier list as well. S tier means I would actually build this. A [music] tier means it's a strong opportunity. B tier means it can work, but there are caveats. C tier means possible, but not my favorite. and F [music] means I think you should stay away from this. By the end, you should know which AI money methods are real and which [music] are not. Let's get started. The fastest way to make money with AI is usually helping a business find customers.

This is why AI lead generation is first on my list. Every business wants more pipeline. They want better prospects, cleaner lists, and more money. That's what we all want. And AI can help with all of that. What this essentially means is you either selling services to businesses where you help them find and acquire clients on X, LinkedIn, email, so on and so forth, or you doing this with your own business in order to increase your own sales.

The important part is that this should not just be scraping 10,000 emails and blasting garbage outreach. That is what I like to call spam. The real value is using AI to understand who the company is, why they might need your help, and what message would make sense for you to send them specifically. For example, my business partner developed a lead generation scoring automation and uploaded an entire video about it that I find very interesting because the core idea is not just that it finds more leads.

It helps our sales team better prioritize who looks like a good fit, who is showing intent, who should get a different message. That is where AI starts becoming super useful. If I were starting this myself, I would pick one single niche or one single offer and one single source of leads. Then I would build a simple system that finds prospects, scores them, gives me the reason they might care, and drafts a first message I can improve.

The trap in this method is thinking volume fixes bad targeting because it doesn't. AI makes it easier to send more bad messages, which means it can only destroy your reputation faster. My verdict on this is a tier. If it is niche specific and tied to a real offer, it can be S tier, but generic AI lead genen is already crowded. Finding the lead is only step one. The next question is, what happens when that lead replies, fills out a form, or lands in your inbox?

Onto method two. A chatbot only makes money if it saves a human from wasting time on the wrong lead. I would call this category entirely chat bots, but I think that word makes it sound a lot smaller than what it actually is. The real business here is lead qualification and routing. Here's what I mean. Someone emails you, someone fills out a form, someone asks a pricing question or replies to a cold email. You typically avoid all those emails because you think they're just a waste of time.

But all of these are actually sales opportunity waiting to be seized. The AI's job here would be to understand what this person wants, qualify them, route them, and help the right human respond faster. We already have versions of this internally. And I have a fun story about this actually. About two months ago, my YouTube channel reached a size where I was receiving a ton of emails from sponsors requesting to work together.

Most of these conversations would go nowhere. So eventually I stopped answering. Then on my way to my vacation, I asked my Hermes agent, "Hey man, can you take care and respond to emails for me? Only message me if you need approval for a dollar figure." I didn't think much of it at the time or honestly think much would happen. Then a few days into my holiday, I decide to check in on Hermes and I see he's asked me for approval on rates for a lead that's interested in our business services.

Hermes was asking me if he could send him a link to our calendar for a meeting and I approved it. Then my team come back to me a few days later and tell me, "Hey, by the way, we just closed the client that you sent us." Me asking Hermes to do that for me took me 5 minutes to set up before my holiday and made us an extra $12,000 in 1 and a half weeks. That is real value. That is much more valuable than a random website bot saying, >> "How can I help you today?" Because businesses do not pay for chat bubbles.

They pay for speed, clarity, and lots and lots of fewer missed opportunities. You could again build something like this for your own business or you could sell it as a full service. If you can guarantee people results and correlate this to money, you're good to go. My verdict for this is a tier. It's not as sexy as AI agents, but it is as close to revenue as you can be. and very easy for a business to understand. Once you can find leads and qualify them, the next level is building the actual AI workers that do jobs inside the business.

Method three, look, businesses do not just wake up wanting an agent. They want a job done without hiring another person. That's what they want. That is why AI agent development is one of the strongest opportunities on this list. The market is moving away from give me a chatbot and it's moving towards give me something that handles this for me every single day. That could be a sales research agent, a reporting agent, customer success agent, an internal operations agent.

The value is not that it is AI. The value is that a job gets done with less human drag. This is also where I think a lot of people will mess up. They will say they can build any agent for any business. That sounds flexible, but it is actually a nightmare. Every client becomes custom. Every workflow becomes messy and every edge case becomes your problem. The better version is to pick one type of agent for one type of business.

For example, a client reporting agent for agencies. that's it. Or a sponsor inbox agent for creators or a lead research agent for B2B teams. The reason this works is simple. If the agent saves a team 10 hours a week, it or prevents missed revenue or helps the business move faster, there is budget for it. The trap is selling the technology instead of the outcome itself because nobody cares that you use the fancy model or framework. if the workflow still breaks every Tuesday.

My verdict for this one is S tier, but only if it is specific. Generic AI agent development is going to become a race to the bottom. [snorts] Specific agents that solve painful business problems, that is the opportunity. But not every AI business has to start with code or agent infrastructure. The next one starts with attention. AI content only makes money when it creates trust, leads or sponsorships. This is where people confuse the tool with the business.

AI writes content is not a business. A business is helping a founder, creator or company turn their expertise into demand. That can mean expos, LinkedIn posts, YouTube scripts, newsletters, lead magnets, sponsorship decks, sales pages. But the content has to do something specific. It has to create trust. It has to create a pipeline. It has to attract sponsors. And it has to move the business you're selling it to forward.

This is the logic behind one of our services, Founderfunnel. The value is not just generating posts. The value is taking messy founder thoughts, sales calls, client wins, product lessons, and turning them into content that sounds like the person and creates opportunities. This can also become a sponsorship business. If you build an AI personal brand and attract the right audience, sponsors will pay to reach that audience.

But again, the audience has to trust you. AI can speed up content production, but it cannot fake trust and authenticity. If I were starting this, I would pick one input and one business outcome. For example, take a founders's weekly voice notes and turn them into posts or their weekly meetings with their teams or their outreach angles or one lead magnet. Then measure whether it creates replies, calls, sponsors or sales.

The trap here is just like you're going to notice with the other ones, it's generic AI copyrightiting. The internet does not need more average posts written faster. It needs sharper content, better content, and a real point of view. My verdict for this one is A tier. It becomes S tier if it has taste or proof and a real business outcome attached to it. Content can create demand. But there is another version of AI content that removes the person completely and that is where I get a little bit more skeptical.

Faceless AI channels, they can work, but this [snorts] is only where AI slop goes to die. I want to be fair here. People are making money with faceless YouTube channels, AI voiceovers, AI avatars, short systems, and essentially content forms. AI makes this easier than ever. You can generate scripts, you can generate voices, images, clips, captions, entire videos with tools that barely existed a few years ago. So yeah, this is a real method.

I would be lying if I said nobody makes money from it. But I would not personally build my whole future on a anonymous AI slop. The moat for it is very weak and so is the brand you build from it and the platform risk. And if the whole advantage is that you can make cheap content fast, someone else can probably make cheap content even faster or even cheaper. The version that can work is not lazy faceless automation. It's a real media operation.

Strong topic selection, strong packaging with real research, good editing, and a reason for people to come back. And at that point, AI is just production leverage. If I were doing this, I would not create random motivational videos of an AI generates a Trump or generic AI listicles. I would pick a specific niche where the viewer wants utility like finance breakdowns, tool comparisons, or educational explainers, then build a repeatable format with actual quality control.

The trap is thinking faceless means effortless. Faceless removes the face, not the need for actual good content. My verdict here is F tier for most people. uh C tier if you have good editorial taste, a strong niche, and a lot of patience. It is possible, but it's not my favorite. The next method has even more hype, and honestly, even more danger. Method six, AI trading. It's useful when it helps you research faster, but it gets quickly dangerous when you treat it like a money printer.

I love playing with AI trading bots. I have had for months where the systems helped me make one or $2,000 a month, but I have also had months where the edge were off, the market changed, and I ended up losing money. That is the part a lot of AI trading content does not want to admit. A trading edge is not a magic spell. It decays with time. other people find it and the thing that worked last month can stop working next month.

The version I like a lot more is AI assisted research. It track markets, monitor smart money wallets, label opportunities as copy candidate, watch list or skip. This is especially interesting in prediction markets like poly market because there are real signals to analyze. But the system should start paper trading first. It should help you make better decisions before it touches real money. As much as I love trading bots, my verdict for this is B tier for most people.

It has high upside and high interest, but it is risky, inconsistent, and not where I would tell beginners to start with their AI journey. Trading is the risky version of AI leverage. But the next method is the boring version. And boring is usually where the reliable money is. We love reliable. The most reliable AI business is often a service people already buy delivered faster with AI. This is one of my favorite categories because it does not require convincing the market to want something new.

The market already buys content. It already buys clips. It already buys sales support, operations, editing, lead generation, whatever. AI just lets you deliver some of these faster, cheaper, or better. Founderfunnel is a great example. The offer is not we use AI. The offer is a business outcome, which is better founder content, more consistent output, faster production, stronger distribution. That distinction matters.

If you sell AI, people compare you to other tools. If you sell an outcome, people compare you to the pain they already have. This is also easier to start than a full software company. You can begin manually. Use AI behind the scenes. Build your workflow. Learn what clients actually want. Then slowly productize the process of delivery. If I were starting from zero, I would pick one painful service category and ask, "What part of delivery can AI make three times faster without lowering the quality?" Then I would package that into a simple monthly offer.

My verdict for this one, this is going to have to be the first S tier. This is one of the best places to start because it connects AI to existing demand. Once a productized service is working, the next question is how you operate it without drowning in the work. That is where AI assistants come in. The best AI assistant is not the one that chats with you. It is the one that keeps the business moving. This is where tools like Hermes Agent and OpenClaw become interesting.

Not because they answer random questions, but because they can run recurring workflows. remember preferences. They can use tools. They can manage task and they can operate across the systems you already use. A real AI chief of staff can do morning briefs, check your emails, access all relevant files, check on clients, help you with content, and so much more. The money here is often indirect, but it is still real. If it saves you hours every week, prevents missed follow-ups, finds opportunities faster, those are all related to value indirectly.

The mistake is treating it like just a random chatbot. Just like you would with chatbt or clot, a chatbot waits for you to ask questions. A chief of staff agent has routines, boundaries, skills, and approval gates. If I were setting this up, I would not start with 50 automations. I would start with one daily workflow. What should you know every morning? What needs a draft? What needs approval? And what should never happen without your approval?

The trap is giving the assistant too much access before you trust the workflow. This is why I recommend you start with read only, then earn autonomy slowly. My verdict for this is a tier for most people, S tier for operators. If you already have a business, clients, content, inboxes, this can become serious leverage. And that leads to our final and most underrated AI money method. Using AI not to sell AI, but to grow the business you already have.

Method nine. The most underrated AI money method is fixing the revenue leaks already existing inside your business. This is the category that does not always make the best thumbnail, but it might be the most valuable one. Every business leaks revenue. Leads do not get followed up with. Clients go quiet. Renewal dates get missed. Churn risk shows up a little late. AI can help with all of that and more. This is where systems like my client pulsebot become interesting.

It scores clients, analyzes clients for churn, tracks unanswered messages, renewal reminders, upsell signals, the list goes on and on and on. This is different from selling AI services to other people. This is using AI inside your own business to make more money from the machine you already have running. And honestly, this is where a lot of operators should start. You do not need a brand new AI business. If your current business is already losing money through slow followup or weak retention or missed opportunities.

If I were applying this, I would answer these three questions. Where do my leads get lost? Where do my clients churn? And where does my team forget to follow up? Then I would build one automation around each one of these. The trap is automating random tasks instead of revenue bottlenecks. Just because something can be automated does not mean it should be. My verdict for this one is S tier. It is not the flashiest method, but it is the most operator native.

If it saves one client, creates one upsell, or prevents one missed deal, it can pay for itself fast. So, if I had to pick one lesson from all nine methods, it would be this. The best AI money opportunities are the ones closest to real demand. And here we have it. Here is how I would rank every single one of them. In S tier, we have AI agent development, AI productized services, and AI revenue automation. In A tier, we have AI lead generation, AI lead qualification, AI content engines, and AI chief of staff agents.

In B tier, we have AI trading bots. In S tier, we didn't end up with anything. And in F tier, we did end up having faceless AI content channels. Look, making money with AI in 2026 is not about finding a secret prompt. There's no secret prompt. Trust me, it is about finding a painful business problem and using AI to solve it faster and packaging the result in a way that people already understand. Don't lose them with AI gibberish.

The closer you are to revenue, the better the opportunity or the easier it will be for you to sell it to clients. This is where AI actually becomes valuable. If I were starting from zero, I would not start with the flashiest method. I would start with one workflow, one niche, one a painful problem and one clear outcome. That is how you avoid the common guru trap. You do not need AI to make money magically. You need AI to make something valuable happen more consistently.

Comment which method you would build first because I'm curious where everyone's head is at. And if you've enjoyed this video, make sure to subscribe because I have a ton more content coming your way. And by the way, the algorithm gods told me you will really enjoy this video next. So maybe I'll see you there. Who knows?