This 22-Year-Old Built TikTok for Mobile Games, and It’s Growing Fast | E2276

Key Concepts

• Nanog: A mobile platform described as "TikTok for casual video games," featuring AI-driven game creation and a feed-based consumption model.
• 1X (Neo): A humanoid robot company focused on general-purpose labor, safety, and physical AI for home and industrial use.
• World Models: AI architectures that simulate spatial and temporal dynamics (3D physics over time) rather than just processing 2D text or image data.
• Embodied Intelligence: The concept that AI learns more effectively when it interacts with the physical world through a body, rather than just training on internet data.
• Teleoperation (Teleop): A method where human experts remotely control robots to perform tasks, providing high-quality data for the AI to learn from.
• Agentic Harnesses: Software frameworks that allow AI to use tools (e.g., generating 3D meshes, pixel art, or sound) to complete complex tasks autonomously.

Nanog: The "TikTok for Games"

Main Topics & Key Points:

• Platform Mechanics: Nanog allows users to play casual games in a vertical, scrollable feed. When a user gets bored, they swipe to the next game, mirroring the "doom-scrolling" behavior of short-form video platforms.
• AI Creation: Users can generate games in 90 seconds using AI prompts. The engine uses Google’s Gemini combined with custom agentic tools to generate assets (3D meshes, 2D pixel art, sound) and code.
• Remixing: A core feature is the ability to "fork" or remix existing games. Users can take a game created by another user and add features (e.g., adding guns to a 3D Flappy Bird) via simple prompts.
• Engagement Data: Since launching in mid-January, the platform has reached 100k users. 20% are "power users" who play over 25 games per session, averaging 21 minutes per session, twice daily.

Business Model & Strategy:

• Interactive Advertising: The founders propose a model where brands (e.g., Domino’s) create interactive, branded games (e.g., an Asteroids-style game where boulders are pizza ingredients) with direct calls to action.
• Creator Economy: They are experimenting with creator currencies and microtransactions, though they emphasize that the current focus is on building a "highly addictive" toolset first.

1X: Humanoid Robotics and Neo

Main Topics & Key Points:

• Evolution from Industrial to Home: 1X began with "Eve," an industrial robot on wheels. They pivoted to "Neo," a humanoid robot designed for the home, because they believe general-purpose labor requires a human-like form factor to navigate human environments.
• Design Philosophy: Neo is designed to be soft, lightweight (66 lbs), and compliant. This ensures safety and allows the robot to interact with the world using human-like dexterity (e.g., using fingers rather than claws).
• The "World Model" Advantage: Unlike Vision-Language Models (VLMs) that treat the world as a series of 2D screenshots, 1X’s world models simulate 3D physics and temporal progression. This allows the robot to "think" ahead and simulate the consequences of its actions before executing them.

Methodologies & Real-World Applications:

• Bootstrapping Intelligence: 1X uses a combination of web-based training data and physical robot data. By having robots in homes, they create a "flywheel" of learning where the robot observes, attempts tasks, and improves via teleoperation feedback.
• Elder Care & Companionship: The founders view Neo not as a replacement for humans or pets, but as a new type of companion that can handle physical labor (laundry, tidying) and provide dignity to the aging population.
• Manufacturing: 1X operates an "all under one roof" facility in San Carlos, handling everything from R&D and AI training to raw material refinement and assembly. This vertical integration allows for rapid iteration.

Key Arguments:

• Safety through Design: By making the robot physically soft and low-energy, 1X argues they can achieve safety through hardware, which is more reliable than relying solely on software constraints.
• The "10,000 Robot" Baseline: The founders estimate that 10,000 robots in the field would provide an influx of data comparable to the current growth rate of YouTube, which they believe is sufficient to achieve significant advancements in general intelligence.

Synthesis and Conclusion

Both Nanog and 1X represent a shift toward generative, interactive, and embodied AI. Nanog is democratizing game development by turning it into a social, short-form content experience, while 1X is attempting to solve the "general labor" bottleneck by creating robots that learn through physical interaction. Both companies emphasize that the "permissionless" nature of their platforms—allowing users to create, remix, and experiment—is the key to rapid innovation and scaling. While Nanog is currently focused on user retention and engagement, 1X is focused on the long-term, capital-intensive challenge of scaling manufacturing and refining world models to achieve true AGI-level physical competence.