Open source AI models have been on the rise, and there are growing concerns that they could soon surpass their closed counterparts. Companies like Google, which have long dominated the AI landscape, are starting to worry about the potential threat that open source models pose to their business. While Google’s large language models are good, they may not be able to keep up with the pace of innovation. Open source models offer faster and more diverse development.
Advantages of Open Source AI Models
According to Luke Sernau, a senior software engineer at Google, “We have no moat. And neither does OpenAI.” Sernau believes that open-source models are faster, more customizable, more private, and pound-for-pound more capable than closed models. While closed models still hold a slight edge in terms of quality, the gap is closing astonishingly quickly. One of the main advantages of open source AI models is the ability for people to iterate on the code. When a problem arises, developers can quickly fix it and push out updates, allowing the model to evolve and improve over time. This is in stark contrast to closed models. They are often restricted in their functionality and require approval before any changes can be made.
The Recent Outpouring of Innovation in the Open-Source Community
The recent outpouring of innovation in the open-source community can be attributed to the release of Meta’s LLaMA model, which was leaked to the public in March of this year. This led to a flurry of ideas and iteration from individuals and institutions around the world. Many of the new ideas came from ordinary people. They were able to train and experiment with the model on their own laptops.
Solving Problems that Closed Models are Still Struggling With
The innovations that have powered open source‘s recent successes directly solve problems that closed models are still struggling with. For example, low rank adaptation, or LoRA, is an incredibly powerful technique that closed models could benefit from. Being able to personalize a language model in a few hours on consumer hardware is a big deal, particularly for aspirations that involve incorporating new and diverse knowledge in near real-time.
The Value of Owning the Ecosystem
Directly competing with open source is a losing proposition for closed models. Who would pay for a closed product with usage restrictions if there is a free, high-quality alternative without them? It’s no surprise that the modern internet runs on open source for a reason. It has some significant advantages that closed models cannot replicate. The value of owning the ecosystem cannot be overstated. Meta, which owns the leaked LLaMA model, has effectively garnered an entire planet’s worth of free labor. Since most open–source innovation is happening on top of their architecture, there is nothing stopping them from directly incorporating it into their products. By owning the platform where innovation happens, Meta cements itself as a thought leader and direction–setter, earning the ability to shape the narrative on ideas that are larger than itself.
Conclusion
In conclusion, the rise of open source AI models poses a significant threat to closed models. Companies like Google need to start paying more attention to what’s happening in the open-source community. By working with these types of models and incorporating them into their products, they can stay competitive and continue to drive innovation in the field.
__
If reading this article on Evotek has piqued your interest, you may be looking for the perfect Vietnamese software development company to help with your next software decision.
Evotek, with their abundance of knowledge and experience around software development and consultancy efforts, is undoubtedly the first port of call.
Get in touch with us today, on +84 (0) 24 6685 0596, or send us an email to [email protected]. Hanoi, Vietnam
