On July 4, Robin Lee, the founder and CEO of Baidu, once again talked about the controversy between open source and closed source big models at the 2024 World Artificial Intelligence Conference and High level Conference on Global Governance of Artificial Intelligence Industry Development Main Forum. He said that the open source big model has value in specific scenarios such as academic research and teaching, but it is not applicable to most application scenarios. "When you are in a fiercely competitive environment and need to make your business more efficient and cost-effective than your peers, a commercial closed source model is the most effective." At the same time, it is emphasized that the focus of large models is still on "rolling applications". "Without applications, just having a basic model is worthless, whether it is open source or closed source.".
"Since this year (2024), many people have confused the concept of model open source and code open source", Robin Lee said, "model open source can only get a bunch of parameters, but also needs to do SFT and security alignment. Even if you get the corresponding source code, you don't know how much proportion and what proportion of data is used to train these parameters. It is impossible for everyone to collect firewood and get these things, and you can't stand on the shoulders of giants for iterative development."
Robin Lee said frankly that under the same parameter scale, the open source model is not as powerful as the closed source model
He acknowledges that open source models have value in some academic research and teaching fields, and can be used to study the working mechanism of large models and form theories. However, open-source models are not suitable for most application scenarios. In a fiercely competitive business environment, if you want to achieve higher business efficiency and lower costs than your peers, a commercial closed source model is the most effective. He took Baidu's practice in novel creation as an example, and after transitioning from open-source models to lightweight models and then to Wenxin Big Model 4.0, the availability and quality of novel generation have greatly improved, allowing online authors to gain wings like tigers. In Robin Lee's view, the focus of the big model is still "volume application". "Without applications, only basic models are available, and no matter open source or closed source, they are worthless."