On the 19th of July, which is this coming Friday, the ByteDance team will make the initial announcement of a wide variety of new artificial intelligence (AI) model technologies.
These technologies include text-to-image and new videos that are similar to those created by Sora.
The announcement will focus mostly on extended videos and high dynamics. With the help of cutting-edge technology, the impacts of Sora could finally be measured.
According to the sources, the AI big model has been established as the most important priority direction for Group P0 within the ByteDance organization.
Additionally, artificial intelligence video model applications are being developed internally by several teams inside Douyin and CapCut.
Thus, these applications are anticipated to be announced in the not-too-distant future.
ByteDance will be hosting an event in Singapore called “ByteDance AI Luminary Talks.” Which will focus on discussing World Models with a group of highly skilled AI technology specialists.
‘Continuous high-dynamic long video generating solutions’ is the topic that will be discussed in the presentation by Zhou Daquan.
ByteDance officially introduced the “Doubao Model” family in May of this year.
This family of models includes general models, role-playing models, voice replication models, and speech recognition models.
ByteDance asserts that the primary Doubao model is priced at only 0.0008 yuan per thousand tokens in the enterprise market.
This model is capable of processing more than 1500 Chinese characters for just 0.8 fen. Which also renders it 99.3 percent more affordable than the standards that are prevalent in the sector.
