🍬 AI Wrappers: 0 to 7 figures in a year

The Opportunity

AI wrappers work. And are making people butt-loads of money. Like, going from 0 to 7 figures within a year.

But not all of them work, many of them flop. Which ones are working and why? What even is an AI wrapper? We get into this below. And at the end there’s some bonus ideas we think could work ;)

Meme of AI wrappers

What is an AI wrapper?

First, what even is an AI wrapper?

Think of an AI wrapper like putting a superhero cape on a cat. You're taking something normal and jazzing it up to make it, super. Practically, it’s a UI that makes requests to an AI model on behalf of a user. The most famous wrapper is OpenAI’s Chat-GPT, which makes it easy to chat with OpenAI’s gpt models.

Many people on the internet are saying these aren’t real products. Maybe they’re right. Or maybe that cat in a batman suit is actually flying between buildings. Let’s see.

Successful Case Studies

First, let’s see how valuable some AI wrappers have become in a short 1-2 years.

Photo AI

Pieter Levels (levelsio on twitter) was the first to popularize the photo avatar projects that seem to be everywhere now. According to his profile, it’s currently making him $55k/month.

Formula Bot

David Bressler asked Chat GPT for excel formulas, and then decided it could work as its own AI app. So he built an AI wrapper around this task and started marketing it on reddit. Now it’s making him $300k/year.

PDF AI

Damon Chen created a web app that allows you to “talk” with pdfs. You upload your pdf, which an AI model will train itself on. Then you can ask it things like “what does this mean?” or “explain this part”. He launched it in May 2023, and within 5 months it was making him $100k a month.

SiteGPT

Bhanu Teja created a chat bot that will train itself on your website data. And then talk to customers for you to get rid of customer service tasks. This is a little more advanced wrapper, but the concept is the same of making calls to AI APIs. He just announced they’re making $100k a month, one year after launch.

Arguments against AI wrappers

Criticism #1: What’s the unique value?

The main criticism is, why wouldn’t a customer just talk to Chat-GPT themself? Why do they need to use your website/SaaS at all?

Well, let’s take Grammarly for example. They analyze text in real time and use AI to improve it.

You could just put that text into Chat-GPT, have it analyzed, and then copy it back over to your site.

Or you could have a custom trained model analyze your text in real time, and offer to change it automatically.

No extra copy/pasting needed. There’s a lot of value in making things easier for users. Which is why Grammarly’s AI tool made them billions.

Criticism #2: you could get wrecked by AI changes

This is true. When Chat-GPT released that you could upload files and talk to them, it wrecked a lot of startups that were doing similar stuff.

However the top products are still popular and making them ridiculous amounts money. For example, Damon Cheng’s PDF AI mentioned earlier. It’s the top result on Google search still and is seeing about the same traffic as it was before the GPT release.

This shows the staying power of these products.

Everything is a wrapper

There’s a strong argument that many successful products are a wrapper of something else. Almost all hosting platforms are wrappers around other hosting platforms. Eg. Render, Heroku etc are just wrappers around AWS.

Many popular payment processors like Paddle and Gumroad are just wrappers around Stripe.

Newsletters are just “wrappers” of info. Nothing I’m saying here you can’t find online. But the value is in how it’s curated and presented (and the great jokes, right??).

The value is in the specialization. If Chat-GPT is a swiss army knife, the AI wrappers are individual tools hyper customized for a specific task.

The key is adding your unique value on top of the AI wrapper. Just regurgitating Chat-GPT outputs will not get it done. Each of these successful wrappers trained a custom model around a specific task. Added safeguards to make sure responses are solid. And then used a pretty UI / UX to attract customers.

Some Ideas For You

Let’s get those entrepreneur wheels turning. Some ideas I’ve seen in the wild already that have gotten lots of attention, some not.

  • Create art based on your Spotify listening history

  • Digital coach. You can niche down to specific industries like sports, therapists, personal trainers etc.

  • Legal / Tax assistant.

  • Generate SEO optimized articles.

  • App that generates excuses for not attending meetings based on your calendar and email tone (serious but not serious).

There’s lots of money to be made in AI wrappers. AI is on track to be a $1.8 Trillion sized market by 2030. And a chunk of the value will be captured by AI wrappers. Happy building friends.