Decentralized AI training is emerging as a transformative force, opening up access to AI models that were previously out of reach for most investors. Training a competitive AI model requires significant financial and technical resources, often in the hundreds of millions, which limits participation to a select few companies. However, new decentralized networks are changing this dynamic by connecting various GPUs worldwide, from high-end hardware to everyday consumer devices. These networks not only facilitate computation but also distribute ownership through tokenization. Contributors receive tokens for providing resources, allowing them to have a stake in the AI models they help create. Recent advancements show that decentralized training is moving from theory to practice. Companies like Prime Intellect have successfully trained models with up to 32 billion parameters, and protocols like Gensyn are proving the feasibility of decentralized reinforcement learning. Unlike traditional models housed in a single data center, these decentralized models are fragmented across the network, which means no single entity owns the entire asset. Participants earn tokens reflecting their contributions, creating a unique alignment between those who provide resources and the AI they help develop. Tokenization introduces an economic structure to these AI models, akin to stocks. Access to models can be traded through tokens that may represent usage rights or a share of revenue generated from queries. This setup creates a new asset class where investors can directly engage with AI models rather than investing in companies. As tokenization gains traction in the financial realm, decentralized AI models fit seamlessly into this trend, offering a digitally native investment opportunity that anyone with internet access can tap into. While many decentralized systems are still in development, the potential for a new market is evident. These networks could become a vital resource for training AI, allowing for the pricing of intelligence itself, not just the companies that create it. The integration of crypto and AI in this manner may redefine how value is assigned in the digital economy.