Use cryptocurrencies to fund open source generative AI

The intersection of generative AI and Web3 has become one of the most active areas of research and development within the crypto community over the past few months. Emerging trends such as decentralized computing, zero-knowledge AI, small base models, decentralized data networks, and AI-centric blockchains aim to provide a native Web3 track for AI workloads.

These trends are technological innovations that attempt to bridge the worlds of Web3 and AI, thus countering the centralized nature of generative AI. While creating technical Bridges with AI is critical to Web3's growth, they are not the only way to integrate these technology trends.

What if the way to integrate Web3 and AI is financial rather than purely technical? As it turns out, the programmable finance and capital formation capabilities of cryptocurrencies may be helpful for one of the biggest challenges facing the generative AI market today. What challenge are we talking about? The funding challenge of open source generative AI.

Open source generative AI needs to succeed

Despite the recent level of innovation in decentralized generative AI, the gap with centralized AI technologies is widening, not narrowing. Many believe that blockchain is the best technological alternative to the centralized AI control of large technology platforms. However, the adoption challenges of decentralized AI platforms are enormous.

Decentralized computing is a clear pillar of decentralized AI, but is practically impractical for pre-trained and fine-tuning workloads that require Gpus to be close to and access data sets that are often located behind corporate firewalls. Zero-knowledge machine learning (ML) is too expensive for large base models and doesn't see any real demand in the market. The decentralized data marketplace needs to overcome the same issues that have prevented the data marketplace from becoming a big technology business.

While decentralized AI strives to overcome these frictions, centralized alternatives are accelerating at a frenzied pace, and the gap between the two is daunting. The only trend that makes people believe that decentralized AI can succeed is the rapid development of open source generative AI.

All decentralized AI trends depend on a healthy open-source generative AI ecosystem, but that ecosystem may not be as healthy as it appears.
Open source generative AI faces a huge funding problem

In the last few years, as an alternative to platforms like OpenAI/Microsoft, Google, or Anthropic, we have witnessed an explosion of open source large-scale generative AI innovation. With the release of the Llama model, Meta became the undisputed champion of open source generative AI. Companies like Mistral are getting billions of dollars in venture capital, enterprise platforms like Databricks or Snowflake are pushing open source models, and more and more open source generative AI is being released every week.

While open source generative AI is gaining momentum, a more detailed analysis reveals a different reality. Open source generative AI is facing a huge funding problem. When it comes to large base models, only large companies like Databricks, Snowflake, Meta or well-funded startups like Mistral can keep up with the performance of large closed models. Most releases from other LABS, such as Databricks and Snowflake, have focused on optimized enterprise workloads, while most recent open source research has focused on complementary technologies rather than new models.

The reason behind this phenomenon can be attributed to the astronomical cost of building large cutting-edge models. Any pre-training cycle for a model with more than 2 billion parameters can cost tens to hundreds of millions of dollars and involve a multi-month process with many failed attempts. These costs are beyond the budget of most university laboratories. What's more, many AI university LABS are funded by large technology companies that are direct beneficiaries of the research results.

Making money from open source has always been difficult, and making money from open source generative AI at AI scale is even harder. As a result, open source generative AI is experiencing a huge funding crisis, which could make the gap between it and the leaders of the AI industry even wider.

Through the programmable finance and capital formation capabilities of cryptocurrencies, we are expected to find a new path to solve the funding challenges of open source generative AI, thereby driving further development of the field and closing the gap with centralized AI systems. This requires us to work together to harness new technologies and drive innovation to meet the challenges of the future.