Open source AI models can't improve in the same decentralized way as software like Linux. While the community can fine-tune and optimize, the primary driver of capability—massive-scale pre-training—requires centralized compute resources that are inherently better suited to commercial funding models.

As the model landscape changes rapidly, AI application companies must operate an internal "model factory." Decagon Labs continuously fine-tunes new open-source models for their specific use cases, creating a system to quickly leverage advancements and maintain a performance edge.

Innovative AI startups are moving beyond proprietary APIs to build defensible businesses. They use open-source models to gain the deep control needed for custom fine-tuning, post-training, and unique deployment methods—capabilities that closed-source vendors do not offer and are essential for differentiation.

History in tech shows that open systems like Linux and Android tend to defeat closed ones. The same dynamic is playing out in AI. Open-source models will likely win long-term because they optimize for widespread adoption and rapid innovation, while closed models focus on maximizing short-term profits within a ring-fenced environment.

Releasing a frontier open-source model successfully is a major operational challenge. It requires tight co-design and coordination between the model lab, hardware vendors, inference engine teams like VLLM, and distribution platforms like Hugging Face to ensure the model is usable and performs well from day one.

The open vs. closed model debate is misguided. Citing AI company Decagon, the speaker explains that open-source is superior for production workloads needing low latency and fine-tuning (90% of their use). Frontier models are better for initial use-case discovery, explaining their current market share in an early AI market.

Contrary to past momentum, the most advanced AI startups are increasingly adopting and fine-tuning open-source models. This shift is driven by the need for cost-effective speed and deep customization as their workloads mature and scale.

Mature AI applications are not static calls to a single large model. They are complex systems of many models that require a continuous "AI loop": tracing performance, identifying areas for improvement (cost, speed, accuracy), and constantly iterating by swapping models, fine-tuning, or refining prompts.

At scale, companies rarely deploy open-source models "off the shelf." Instead, virtually all production workloads involve custom modifications. This can be post-training with proprietary data to improve quality or compiling and quantizing the model to enhance performance and reduce cost.