Now, the commercialization of large models is once again on the table.
One fact is that the current large model training needs strong computing power support, especially the model with a large number of parameters, which is extremely expensive. Large model training requires huge amounts of money to support.
After investing a huge amount of money, enterprises on the one hand hope to commercialize as soon as possible to solve the problem of follow-up research and development funds, on the other hand also hope to achieve the purpose of making money through commercialization.
It is inevitable that some safety and ethical issues will be put aside for the time being, and it is more true that there is not much thinking about the development path after the rapid commercialization of large models, so we can see that many large models are basically just a taste, and finally lead to the contradiction between commercialization and non-profit.
First, the commercialization of large models
But in spite of this, the commercialization of domestic large models is still in the initial stage, and the commercialization process faces many challenges.
First, the development and application of large models requires a large investment of money and time, and the return is often difficult to predict. This has led many companies to hesitate in the commercialization process and miss market opportunities.
Second, the ethical and safety issues of large models also put pressure on commercialization. At present, the application needs of most enterprises are mainly concentrated in the fields of intelligent customer service, intelligent recommendation, intelligent marketing, and other fields of application are still in the exploration stage. This makes the commercialization process of large models relatively slow and difficult to achieve large-scale development.
What is more noteworthy is that although China has made significant progress in the field of artificial intelligence, there is still a certain gap in domestic large-model technology compared with the international leading level. This makes domestic enterprises at a disadvantage in the international market competition, and it is difficult to extend to the sea and cross-border direction.
The commercialization of large models should solve the problem of allowing enterprises and users to understand the principle less, and use the results more simply and directly, so that users can return to the value and solve their own business problems. In other words, it is the "integrated black box model" of the large model.
As a result, some of today's business models are a gathering place for big-model racetrack players and entrepreneurs.
Second, what is the difficulty of commercialization?
Overall, the path to commercialization of large models is not the best, but the direction is clear. But clarity doesn't mean landing. For the players of the domestic large model track, they still face many internal and external challenges.
In general, the commercialization of the global large model industry is still in the early stage of exploration.
On the one hand, although research and development institutions have been quite mature in terms of large-model technology, they are not familiar enough with the landing scene and have not yet formed a perfect commercial model. Therefore, they need to work with downstream scenario enterprises to jointly build large model business models.
On the other hand, most downstream scenario enterprises have not yet formed the basic concepts and cognition of large models, and they lack the computing power required to support model fine-tuning, as well as the human resources and technical strength required to customize and secondary develop models.
In general, although the path of commercialization of large models is relatively clear, and domestic manufacturers are actively exploring it, the commercialization of large models cannot be limited to the exploration of business models, but more to solve the underlying problems of the development of large models.
