I quit my job to make $6M/year with AI apps

Key Concepts

• Mimesis: Copying what works and adding your own spin.
• Viral Content Creation Loop: Users create content within the app, share it on other platforms, driving user acquisition.
• Stupid Simple UI: Extremely easy-to-use interface, often with minimal screens.
• Algorithmic Awareness: Understanding and leveraging the algorithms of platforms like TikTok and Instagram.
• Inference Efficiency: Optimizing AI model performance for speed and cost-effectiveness.
• Peer-to-Peer Computing: Utilizing user devices for distributed computing tasks.
• Popcorn Mobile Apps: Simple, entertaining apps designed for mass appeal and quick monetization.

Mimicking Success: The Core Strategy

The speaker, Ben, emphasizes a core strategy of mimesis: identifying what's already working, copying the fundamental principles, and adding a unique twist. He attributes his success in various domains, including mobile app development and viral content creation, to this approach.

• Example: Learning to play League of Legends by watching better players and adapting their strategies.

Wombo: From Meme to Millions of Downloads

Ben details the creation of Wombo, an app that animates selfies to sing songs, as a prime example of mimesis in action.

• Identifying Signals:
  • Reface: A face-swapping app that achieved massive success and venture capital funding. Ben understood the underlying technology and simple design.
  • PewDiePie Video: Showcased the "First Order Motion Model," an open-source AI model that animated faces, going viral within technical communities.
• The "Aha" Moment: Recognizing the potential to create a user-friendly app based on the First Order Motion Model, similar to Reface but with a different AI technology.
• Key Features:
  • Stupid Simple UI: Only four screens: selfie input, song selection, loading screen, and output.
  • Viral Content Creation Loop: Users easily create shareable content, driving organic growth.
  • Strategic Song Selection: Choosing recognizable and meme-worthy songs to enhance virality.
• Results: Wombo achieved over 100 million downloads with minimal marketing spend, relying on the viral nature of the content.

Dream: Replicating the Formula for Image Generation

Following Wombo's success, Ben's team applied the same principles to image generation with the app Dream.

• Identifying Signals:
  • Early viral examples of image generation using a combination of VQGAN and CLIP models within AI research circles.
• Key Features:
  • Simple Input: Text prompt for image generation.
  • Pre-processing: Adding stylistic keywords to user prompts to improve image quality.
  • Fast and Free Inference: Ensuring quick generation times and no upfront cost for users.
• Results: Dream also became a number one app, demonstrating the replicability of the mimesis strategy.

Monetization Strategies

Ben discusses the monetization strategies employed after initially focusing solely on user acquisition.

• Copying the Best: Observing and emulating successful app monetization techniques, particularly those used by Nikita Beer.
• Subscription Model: Offering premium features like generating multiple images at once and access to exclusive styles.
• Ads: Implementing advertising within the app.
• Pricing: Experimenting with different subscription tiers, starting cheap and increasing prices over time.

W.ai: Decentralized Computing and the Future

Ben's current focus is on W.ai, a platform for decentralized computing.

• Motivation: Driven by the need for inference efficiency and the potential of on-device computation.
• Concept: Allowing users to contribute their device's computing power to the network in exchange for rewards.
• Inspiration: Studying existing decentralized AI and computing projects like BitTensor, Render, and Golem.
• Vision: Creating an "Airbnb" for computing power, enabling users to monetize their idle devices.

Creating Viral AI Apps: A Step-by-Step Guide

Ben outlines a practical approach for creating successful AI-powered mobile apps:

1. Landscape Analysis: Thoroughly study existing apps in the target market, paying attention to every detail of their design and functionality.
2. Content Trend Analysis: Identify viral content formats and themes on platforms like TikTok and Instagram.
   • Example: The "pregnant cat" meme, which combines recognizable celebrities, controversial themes, and relatable concepts.
3. Productize Viral Content: Create an app that simplifies the process of creating similar viral content.
   • Example: An app that turns any photo into a "babyfied" image and animates it, abstracting away the complexity of using tools like ChatGPT and Hedra.
4. Focus on Simplicity: Prioritize a "stupid simple" user interface to maximize accessibility and virality.
5. Embrace the Viral Loop: Design the app to encourage users to share their creations, driving organic growth.

Key Quotes

• "It's actually pretty [ __ ] easy if you uh don't overthink it too much."
• "If something is going viral inside of like some technical circle, but it's still it's still difficult for a lay person to use, then uh if you can help the lay person get access to that same technology and make it easier for them, that's typically like a really good recipe for verality."
• "Just copy people who are better than you."
• "If you make an app and it goes viral, even if it goes viral once and people never use it again, like that's a eight, nine, potentially nine figure opportunity."
• "People are glued to their [ __ ] phones and they want to be entertained and they want utility. So if you can be useful or entertaining, then you can make an app."

Technical Terms

• First Order Motion Model: An AI model used to animate faces based on a driving video.
• VQGAN + CLIP: A combination of AI models used for generating dreamlike images.
• Inference: The process of running an AI model to generate an output.
• Inference Efficiency: Optimizing the speed and cost of running AI models.
• Generative AI: AI models capable of generating new content, such as images, text, and audio.

Synthesis/Conclusion

The key takeaway is that creating successful mobile apps, especially in the AI space, doesn't require reinventing the wheel. By applying the principle of mimesis – copying what works and adding a unique spin – and focusing on simplicity, virality, and efficient monetization, developers can create "popcorn mobile apps" that achieve massive downloads and generate significant revenue. The current landscape offers infinite opportunities for AI-powered apps that entertain and provide utility, and venture capital is not necessarily required to achieve success.