Glory despises the online large-scale model track! ! !
The latest news on June 30, 2023 , the MWC conference is currently in full swing in Shanghai, Honor CEO Zhao Ming delivered a keynote speech on the future evolution of smartphones.
During the group interview, Zhao Ming mentioned Honor's development ideas in AI. The appearance of the large model last year made our eyes shine, and made the whole company and Glory team think deeply. When discussing AI, Glory defines AI as an intelligent life that can evolve. In fact, we have given such a concept since December 2016. It can grow with you as you use it. is constantly evolving. This is the train of thought and direction we adhere to at this stage.
Zhao Ming also said that everyone is currently using a general-purpose large-scale network model. No matter how good the field is, it is not a dish for Glory, and it is not our strategic main channel at all. We are still a consumer electronics company. Models, Google big models, Wenxin Yiyan compete in various fields, your genes are not like this.
No matter how good and hot these things are, they are not the company's core business direction and strategy, but we will follow up, learn, think, and learn from others internally.
In addition, Honor will focus on the development of end-side AI capabilities and large models. The big model on the network side is to understand the entire human knowledge base, and the more it learns, the smarter it becomes. Our (end-to-side big model) philosophy is that the more it understands you, the more it can help you. Because objectively speaking, what does the big network model have to do with me? It doesn’t know what time I go to bed or what I like to eat. It can’t solve my needs today, and it can’t gain insight into me.
The end-side large model not only understands you better and can continue to evolve and evolve, it should also allow you to better interact with the network large model and make you more efficient.
In the future, Honor will cooperate with Internet companies. Now many companies and enterprises are also investing in the field of large-scale models, such as many vertical domain models and general large-scale models, covering different fields. The cooperation between Honor and these companies is definitely an inevitable choice in the future.
At the same time, interacting with large models on the network side has also become critical. For example, users will ask questions, what is the potential question he wants to ask? How to improve the efficiency of interaction? The end-side large model not only understands you better and can continue to evolve and evolve, it should also allow you to better interact with the network large model and make you more efficient. From this perspective, collaboration and cooperation between the device side and the network side are inevitable in the future, and each has its own value.
