Weibo Corporation (NASDAQ:WB) Q3 2023 Earnings Call Transcript

First one is still we have to focus continuously on the accumulation and also the collection of data by using and applying AIGC, so that it is going to help those bloggers in generating the content in a more efficient and higher quality way. So, for instance, the contents generated by AIGC have more user stickiness and also interactive rates and very efficient as well. So in Q3, so we have been actively applying the LLM, a large language model on different verticals to increase the efficiency of the content generation. Let me give you example. So in Q3, in the area of the fashion and cosmetics and also in the gaming industry, on a daily basis, AIGC is facilitating the blogger in generating over 5,000 pieces of information a day. And also, if you’re comparing the content generated by AIGC versus naturally or written by the bloggers themselves, you can see a 50% better interactive rate and also user stickiness for the AIGC-generated content.

So we, of course, plan to apply the AIGC well in the beauty and cosmetics vertical, gaming vertical and also the digital and electronics products area. And next year, we are thinking about rolling this technology out to the rest of the other verticals. And also last but not least, we are also having some innovative trials and experiments, for instance, using the AI to learn the characteristics and also the features or the languages that those big Wes that always use to generate an AI chatbot or some of the virtual characters. So, we did that in Q3 to actually really using AIGC to learn from those important KOLs and influencers, learning the way that they talk and also the style of their interaction, so that the AI chatbot was interacting with the fans.

So in August, for instance, we had experimented this particular technology upon 4 to 5 KOLs and also on a daily basis, generating tens of thousands of interactions together with the friends — with the fans and followers using AI chatbot feature. But of course, at the current moment, we are not opening this to the rest of the ecosystem of the all big Wes, but still, we are thinking about doing that in the future still. And nowadays, we are trying to perfect the whole system and also trying to really find out the final effectiveness of this application.

Operator: Thank you. This concludes today’s conference call. Thank you for participating. You may now disconnect.