I Spent 30 Days Making Money With AI. Here’s What Worked

Every week there is a new batch of YouTube videos promising that AI will make you rich if you just follow five simple steps. Most of them leave out the part where you need to actually know something before those five steps do anything useful. I spent the last 30 days putting AI to work in real ways, the kind that brought in real money, and the results were not what the guru crowd would have you believe.

What actually worked was not a magic prompt or some secret tool nobody has heard of. What worked was pairing AI with knowledge I already had, problems I had already lived through, and audiences I had already been building. This post breaks down exactly how each of the five methods worked, what tools I used, and what you would need to bring to the table to get similar results.

What You’ll Walk Out With

The five specific AI methods I used to generate income over 30 days
The exact tools behind each method, including proofinttel.platformproof.com and notes.platformproof.com
Why your existing knowledge is the most important ingredient AI needs to work
How to use the YouTube Studio trends tab to find video topics that are already gaining traction
How to turn real problems you have lived through into digital products people will pay for
A simple lead magnet system that builds your mailing list without any extra daily work
A free tool to figure out which method fits your specific skills and experience at finder.platformproof.com

Why Most AI Money Advice Fails Before You Even Start

There is a phrase you hear constantly in these conversations: making money with AI is like shooting fish in a barrel. That line is designed to make the opportunity sound so obvious that you feel foolish for not already doing it. The problem is that shooting fish in a barrel still requires knowing how to shoot. AI will enhance what you already know. It will not turn an average marketer into a superstar marketer without the work, and it will not turn someone with no background into an overnight success story either.

The biggest mistake people make is walking up to ChatGPT or Claude and typing something like “make me a million dollars.” The tool gives them a general idea, they cannot tell whether that idea is any good, they cannot push deeper into the details because they lack the domain knowledge to ask the right follow-up questions, and when they launch something and it fails they say AI does not work. The real issue is that they brought nothing to the table except hope.

Every one of the five methods below worked specifically because I had real experience in that area. I knew content creation. I had personally lived through the problem of tracking my mother’s medical appointments across multiple family members. I knew what kinds of digital products my audience was already buying. AI made the execution faster and the output sharper. It did not replace judgment, experience, or knowing your market. It amplified all three of those things once they were already in place.

Method 1: AI-Powered Content Ideas That Drive Real Sales

The first method I used was not some clever new product launch. It was getting more of the right attention on content I was already creating, and then selling products to the people who showed up because of that content.

I built a tool called proofinttel.platformproof.com that generates content ideas based on whatever your niche happens to be. You put in what your channel or platform is about and it surfaces topics worth exploring. But the bigger discovery during these 30 days was using AI alongside the YouTube Studio analytics dashboard in a way I had not tried before.

Inside YouTube Studio there is an analytics tab, and inside that tab there is a trends section. When you click on it, YouTube shows you which topics are currently performing at a high level, which are at a medium level, and which are not gaining traction. I used that information consistently over about two weeks. The most recent videos I published based on those insights started getting more views, more engagement, and more interaction than the older content. Part of that was YouTube adjusting how visible view counts are displayed, but the more important factor was that the topics were actually resonating with what people wanted to watch right now.

Getting more views only pays off if you have something to sell behind the attention. The content I was producing drove people toward my digital products and services. AI helped me understand what my market wanted to watch, which meant more of the right people found me. More of the right people meant more sales downstream. That is the chain you are building when you use AI for content strategy: more relevant content leads to more targeted attention, and more targeted attention leads to more buyers.

The important note here is that I have spent years building content and I already have a solid sense of what tends to work in my space. AI gave me better data to confirm or challenge those instincts. If you are brand new to creating content in a niche, AI can still help, but you will need to evaluate its suggestions with less context than someone who has been in the game for a while. Start creating, build your instincts, and then layer AI on top of that foundation. Do not skip the foundation.

Method 2: Digital Binders Built From Problems You Have Actually Lived Through

The second method produced some of the most direct results over the 30 days. I built and sold two digital products: a pet care binder and a parent patient care tracker. Both came directly from problems I had personally experienced. Neither one came from a brainstorming session with an AI tool or a list of trending product ideas I found online.

The parent patient care tracker started when my mother moved in with me. When I took her to appointments and my sister took her to different appointments on other days, there was no clean, easy way to share information between us. I am not in healthcare. I speak a completely different language than the healthcare professionals in my family. My sister knows which questions to ask the doctors. She knows which medications interact with which conditions. I frankly do not know any of that. What we needed was something that made the whole coordination process dummy-proof so that I could show up to an appointment and actually be useful.

The product I built was a web app where family members could log in and share information about medication schedules, upcoming appointments, and doctor visit notes, all in one place without a stream of text messages flying back and forth between people who were already exhausted and stressed. AI helped me write the ad copy, create the images for the Facebook ads, and refine the messaging. All of that happened based on my direction, not based on AI guessing what someone might want. I was the one who knew the problem deeply. AI was the one who helped me build the solution quickly and market it without needing a big team behind me.

The pet care binder came from an equally specific frustration. We have a dog, and the process of switching veterinarians is a nightmare. The old vet does not enthusiastically share your records because they know you are leaving. The new vet asks you a long list of questions about vaccination history, current medications, and past procedures, and if you have not kept meticulous records you are standing there with nothing. And if you ever need to take your dog to an emergency vet at 3 in the morning, your regular vet is closed and the emergency clinic typically does not have access to your regular vet’s records anyway. A web-based binder you can open on your phone solves all of that in one place.

Both products sold through Facebook ads. AI wrote the ad variations based on the direction I gave it. I made small tweaks, ran tests, adapted the messaging based on what the data showed, and kept going. The key insight is that both products solved a specific, named, lived problem. That is where products that actually sell come from. Not from abstract ideation sessions with an AI tool, but from paying close attention to the friction in your own life and in the lives of people around you.

Method 3: Custom Dashboards That Replace Spreadsheets Nobody Wants to Touch

The third method is one of the most interesting for people who want to sell something without needing an existing audience. Most people do not want to deal with Excel spreadsheets. They want to see their numbers in a visual format, with graphs and automatic updates and clean tables, without having to build or maintain anything technical themselves.

You can build dashboards similar to Notion templates that connect to real accounts and update automatically. As a concrete example, you could connect PayPal to a dashboard so that the person’s transaction data shows up as graphs, summaries, and tables without them ever touching a spreadsheet. You could build this once and deploy it for hundreds of users, each with their own account connected, each seeing their own private data. One build, many customers.

The same rule that applies to the binders applies here: the best products in this category start with a problem you have personally faced and solved. You do not need to invent a new category. You need to find a specific thing that caused you real friction, solve it cleanly, and then sell that solution to other people who are dealing with the same friction. Once you have solved it for yourself, you have already done most of the product research.

Not sure which of these five methods fits your background and experience?

Answer a few questions and get a specific recommendation tailored to your situation at finder.platformproof.com.

Method 4: AI-Generated Lead Magnets That Build Your List While You Work on Something Else

This one surprised me with how well it worked, mostly because it required almost no additional effort once the system was in place. The notes and worksheets that accompany every video I make are produced entirely by AI. I do not read them before they go out. The process runs on its own.

Here is exactly how it works. Once I finish a video, I take the transcript and upload it to Claude. Claude turns that raw transcript into structured notes and a worksheet that complements the video content. I put those documents on notes.platformproof.com and offer them as free downloads. People who want the notes give me their email address. They get the notes. I get a subscriber who has already shown they care about that specific topic because they watched the video and wanted to go deeper.

I got this idea from watching a channel called Affiliate Marketing Dude. He has notes available for all of his videos and from what I can tell he uses a virtual assistant to produce them. I replaced the virtual assistant with AI. The cost is essentially zero. The ongoing time investment after the system is configured is also essentially zero. But the result is that I am continuously adding value to the people in my audience and continuously adding subscribers to a mailing list I can sell to later.

This system works because AI can process a raw transcript and produce something clean and useful without me reviewing every word. That reliability only happens because I have spent time giving AI the context it needs to understand how I think and communicate. I upload my transcripts from workshops, from courses I have taken, from past videos, and from other long-form content. AI sees how I talk, the phrases I repeat, the way I structure an argument. It learns to produce output that sounds like me rather than sounding like a generic content template. The more you feed it, the more accurate that mimicry becomes.

If you already have a YouTube channel, a podcast, or any kind of content you are producing consistently, this system is free to start today. Take a transcript, upload it to Claude or ChatGPT, ask it to produce a one-page summary and a short worksheet, drop those files on a simple landing page with an email capture form, and you are building a list. The audience grows while you focus on making the next piece of content rather than spending hours on back-end materials.

Method 5: A Central Hub That Lets All Your Products Work Together

The fifth method is less about AI doing something technically impressive and more about using AI to build something that should have existed long before it did. I needed a central place where all of my digital products and services lived together in one visible, organized location that any visitor could explore.

Before building the hub, I had products and workshops scattered across different links and different platforms. Someone who found one video might have no idea I had a workshop teaching people how to create a digital product from scratch in three days. Someone who found the workshop might have no idea about the binders. There was no clear path from one offering to the next, and that invisible barrier cost sales every day.

AI helped me build something similar to a School-style community site where I can run an online community, sell digital products and services, host past workshops, and give people a single place to come back to. You can build something like this using Claude, Codex, or several other tools depending on your technical comfort level. The result is a hub where everything is visible and people do not have to hunt through your entire content history to find what you offer.

This particular method pays off most once you already have multiple products or services worth collecting in one place. If you are just starting out with one product, a full hub is premature. Build the first product, sell it, understand what your buyers actually want next, and build the second thing. Once you have a few things that are selling, the hub becomes the infrastructure that makes all of them easier to find and buy.

Bonus: Using ChatGPT to Price and Negotiate on Facebook Marketplace

This method came up during the 30 days almost by accident and it is worth including even though it operates at a smaller scale than the others. Like most people, I have old items sitting around the house that have value but that I never get around to selling because pricing them and writing a good listing feels like more work than it is worth.

I started taking photos of those items and uploading them to ChatGPT with a simple question: what price should I list this for on Facebook Marketplace? It gives you a starting point based on condition, category, and current demand. You can also use it on the buyer’s side: if someone sends you a price you think is too high, you can paste their message into ChatGPT and ask it to draft a polite counter-offer. This will not replace any of the five main methods above, but it is a legitimate use of AI that has a real dollar value and takes less than five minutes per item.

Honest Drawbacks: What AI Still Cannot Do For You

None of this is as frictionless as the YouTube thumbnails make it look. There are three real constraints worth naming clearly before you walk away thinking you just need to open a chat window and start generating income.

First, AI is only as useful as the direction you give it. If you feed it vague instructions and no context about your niche, your audience, or your experience, you will get generic output that does not convert. Every result described above came from giving AI specific, informed direction based on real knowledge. That means you have to do the thinking first. AI handles the execution once the thinking is done. It does not replace the thinking.

Second, every method that worked had something to sell attached to it. Content ideas only generate income if you have a product behind the attention. Binders only sell if you have a distribution channel in place. Lead magnets only turn into revenue if you have something to offer the list once it is built. If you are starting from scratch with no product, no audience, and no experience in a particular area, you will need to build at least one of those three things before AI becomes a meaningful income tool for you.

Third, there is no guaranteed timeline. I am not promising you will see a dollar from any of this by a specific date. What I can say with confidence is that the people who will definitely not make money with AI are the ones who give it a vague command and walk away expecting income to appear. The people with a realistic shot are the ones who bring actual knowledge, an actual problem they want to solve, and the willingness to test and adjust when the first attempt does not perform the way they hoped.

Find Your X

If you read through all five methods and you are not sure which one fits your specific situation, that is exactly the problem the Platform Proof Finder was built to address. You answer a few short questions about your background, what you already know, and what kind of income you are working toward, and you get a specific recommendation rather than a generic list of options. Go to finder.platformproof.com to get your result.

Frequently Asked Questions

Do I need technical skills to use AI to make money?

Not for most of the methods covered here. Writing ad copy, generating content ideas, creating notes and worksheets, and using ChatGPT to price Marketplace listings all happen through a plain text conversation. Building dashboards or web apps requires more technical knowledge, but even those can be built with AI doing a significant portion of the coding work if you are willing to learn as you go and ask very specific questions about what you want the tool to produce.

Which AI tools did you actually use during the 30 days?

Claude for converting video transcripts into notes and worksheets. ChatGPT for Facebook Marketplace pricing recommendations, ad copy drafts, and ad image direction. proofinttel.platformproof.com for structured content idea generation tied to a specific niche. The YouTube Studio analytics trends tab was the most practical discovery tool for identifying which topics were gaining momentum in my space right now rather than six months ago.

Can these methods work if I have no existing audience?

Yes, for some of them. The pet care binder and the patient care tracker both sold through Facebook ads with no existing audience required. If you have a product that solves a real problem and you are willing to spend some money on targeted ads, you do not need followers to make your first sales. The lead magnet and content strategy methods work better if you already have some kind of platform, but you can build the platform and run the lead magnet system at the same time so that one feeds the other.

How long does it take to build and launch a digital binder or dashboard product?

With AI handling the build and the initial marketing copy, you can have a working version in a matter of days rather than months. The pet care binder and the patient care tracker both came together relatively quickly. The larger time investment came from testing the Facebook ads, reading the performance data, and adjusting the messaging until the conversion rate made the ads profitable. Budget two to four weeks from idea to a product that is actually selling at a break-even point or better.

What makes a strong idea for a digital product built with AI?

It starts with a problem you have personally experienced and then solved for yourself. Not a problem you read about in a blog post or think other people might be dealing with in the abstract. A problem that caused you real friction, that you researched your way out of, and that you eventually fixed. The pet care binder came from the lived experience of switching vets and realizing medical records do not transfer easily. The patient care tracker came from the real stress of coordinating a parent’s medical care across multiple family members with different schedules and different levels of medical knowledge. That personal depth gives you enough specific detail to build something actually useful rather than something generically adequate.

How do I train AI to write in my voice rather than sounding generic?

Upload your transcripts. If you make videos, take the auto-generated captions and paste them into Claude or ChatGPT at the start of a session. If you have completed courses or run workshops, upload those transcripts too. The more examples you give it of how you actually speak and write, the better it gets at recognizing your recurring phrases, the rhythm of your arguments, and the way you explain a concept to a specific type of person. It will not be a perfect reproduction, but it gets close enough that the output feels like yours rather than like a generic content template.

Is Facebook advertising worth the cost for a low-priced digital product?

It depends heavily on your cost per click and your conversion rate for that specific product at that specific price point. Low-priced products need a high volume of sales to cover ad spend, which means your targeting and your copy have to be tight from the beginning. The advantage is that a lower price reduces the barrier to purchase, which often means you collect performance data faster because more people are willing to try it. Many sellers use a low-priced front-end offer to bring buyers in and then sell something at a higher price point on the back end to the people who bought the first product.

What is the single most important thing to get right before trying to make money with AI?

Know what you are selling and who you are selling it to before you open the AI tool. The people who see results from these methods come into the conversation with a specific problem they understand, a specific audience they know something about, and some existing experience in that space. The people who struggle are the ones hoping AI will identify what to sell, who to sell it to, and how to persuade those people to buy, all without them having to bring any existing knowledge to the table. That is asking too much of a tool that depends entirely on your input to produce anything worth using.

Sources

Platform Proof AI content idea generator: proofinttel.platformproof.com
Platform Proof notes and worksheets hub: notes.platformproof.com
YouTube Studio Analytics, Trends tab (found inside the Analytics section of YouTube Studio)
Affiliate Marketing Dude YouTube channel (inspiration for the transcript-to-notes lead magnet system)
Original video: “I Spent 30 Days Making Money With AI. Here’s What Worked” by Alston Godbolt, Platform Proof (YouTube, https://youtu.be/zy5WDOsbMro)