Kunlun Wanwei CEO Fang Han: Generative AI data has encountered bottlenecks, and it is recommended to build industry-specific large models

Kunlun Wanwei CEO Fang Han: Generative AI data has encountered bottlenecks, and it is recommended to build industry-specific large models

According to reports on May 19, Fang Han, CEO of Kunlun Wanwei Company, recently elaborated on his views on generative artificial intelligence on the official public account. Fang Han pointed out that the bottleneck of generative AI lies in data. The data threshold is still very high. The business model of large-scale AI models in the future can only be realized by reducing costs and improving efficiency. For the industry, it is necessary to embrace large models as soon as possible and make proprietary large models for its own industry, and specifically stated that "it is not recommended that the game industry use AIGC, and more attention should be paid to how to use AIGC technology to improve gameplay." In addition, he is optimistic about the prospects of multi-modality in large model engineering.

Fang Han is also optimistic about the prospects of multi-modality in large model engineering. Integrated multimodal technology can handle various types of data, including audio, text, and images, further enhancing the capabilities of large models. The research organization IDTechEx recently released a report looking forward to the AI ​​chip market trend from 2023 to 2033. It is predicted that the AI ​​chip market will grow to US$257.6 billion by 2033, and large model technology will also become an important growth driver.

The application of AI technology has gradually penetrated into various industries. Among them, the global game market will reach RMB 1,110.76 billion in 2022. The global mobile game market size is approximately RMB 594.519 billion. In the game industry, it includes NPC behavior simulation, game scene generation, screen optimization and other aspects. In these applications, generative AI technology plays an important role in promoting.

For example, in "The Witcher 3", the artificial intelligence of NPC characters can respond accordingly according to the player's behavior, making the game experience more realistic. In "Red Dead Redemption 2", the use of generative AI technology has greatly improved the details and image quality of the game scenes.

Game developers can produce high-quality game content faster by using generative AI technology, thereby reducing production costs and increasing competitiveness.

In the future development of the game industry, AI technology will play an increasingly important role. In the future competition of game developers, how to skillfully apply AI technology to improve game experience, improve efficiency and reduce costs will become their core competitiveness.

The above is the detailed content of Kunlun Wanwei CEO Fang Han: Generative AI data has encountered bottlenecks, and it is recommended to build industry-specific large models. For more information, please follow other related articles on the PHP Chinese website!