SenseTime, a pioneering Chinese artificial intelligence firm, has just released their large-scale model system SenseNova, complete with a chatbot tool known as SenseChat. This real-time model contains a combined 100 billion parameters, allowing it to complete tasks such as storytelling, code-writing and medical advice. SenseChat could potentially stand in stark contrast to ChatGPT, which contains 175 billion parameters.
In addition to the tool, the SenseNova model includes components for text-to-image generation, and even an avatar video generator, making it a comprehensive suite for custom applications. SenseTime, unlike some other major players in the field such as Baidu with ERNIE Bot, and Alibaba and their Tongyi Qianwen chatbot, is currently not offering its AI-powered products for public testing, instead, restricting access to government and corporate clients.
The CEO of SenseTime, Xu Li, unveiled the model at Monday’s showcase, while the founder Wang Xiaogang revealed the architecture and discussed the power of the model. Hopefully, in the future, they will open up access to the public, allowing the SenseNova suite to become available to all.
