So, we will stick on this route and making sure that all of our game will sustain a long life operation cycle.
Juliet Yang: Operator, next question, please. Thank you.
Operator: We are now taking the next question. Please standby. The next question is from Lincoln Kong from Goldman Sachs. Please go ahead. Your line is open.
Lincoln Kong: [Foreign Language] So, my question is about AIGC’s application to Bilibili. So, from past Investor Day, we already seen some development or strategy showcase in terms of AIGC speeding up the production efficiency, the marketing promotion tools as well as the AI-integrated search. Could management elaborate more around how we think about AGIC could benefit Bilibili?
Rui Chen: [Foreign Language] [Interpreted] We believe this is breakthrough technology of — AIGC, what can bring to the industry. The way we look at it, it equals to the mobile Internet that happened 10 years ago. And this will not only benefit Bilibili, it will actually benefit everyone. So, as for Bilibili, our core business is video and ACG, and it has been very clear that AIGC can enable an increased efficiency in a large amount on both fronts, for example, on content production as well as content recommendation. Also in terms — based on the LLM, we can increase a lot of the efficiency of managing our community by understanding what the text means, the content means within our community. And we believe this technology, AIGC, has multifaced applications and scenarios to help us to improve all around user experience.
As for the application scenario, actually we have already conducted some extended research and development, and we believe there’s a lot of scenarios that we can leverage. For example, the prepared remarks that spoken by Sam just now before this Q&A session, actually, it was not actually conducted by Sam. It’s powered by our self-developed AI tool, text-to-speech, TTS, tool. And similar products like this, we have already starting to research and develop and apply it in our various scenarios. And there are other examples. For example, this year’s Bilibili Macro Link, our live virtual host, 22 and 33, their voice was also generated by the same tool, TTS AI tool that Bilibili self-developed. And another example would be knowledge-related video creation, as many knowledge content creators, they generally write the text and they record their voice.
With the help of this AI tool, they can — it can largely improve its production efficiency. As for the LLM, even though we haven’t marketed quite aggressively, but actually, we have already started to apply LLM in various daily operations. Another example would be on the content auditing side. In the past, we have to spend a lot of human resource and power in managing the content with the help of large language model, because it can understand the voice, the text, the meaning and all of the implied meanings behind the text, it can drastically increase the efficiency of how our content auditing process the content. And another application scenario would be enhanced search, because Bilibili has a vast content library and each video carries a lot of scatter information and we can leverage the LLM to drastically enhance the search efficiency.
We’re actually beta testing a new enhanced search that by just typing question in our search box, our large language model can summarize and answer questions. Very valid example would be, with the recent version upgrade of Genshin Impact, people can just simply ask which card is worth drawing and which character is worth spending time to improve, and the large language model can summarize the valid information and just provide to the user versus in the past, they have to watch several videos to summarize that content information. Now, they — with the just stroke of hand, they can get the correct and fast answer. Another function upgrade that we are testing is AI Assistant. This will accompany users when they are watching a video, help them to summarize and extract useful information.
