Robin Li, CEO of Chinese language tech firm Baidu, stated on Sep. 5 that greater than 70 synthetic intelligence (AI) fashions have been launched within the nation, in accordance with a report from Reuters.
Li commented that the over 70 AI language fashions which have been launched have been created with over 1 billion parameters, which technically qualifies them as giant language fashions (LLMs).
At the moment, among the greatest LLMs in the marketplace embody Meta’s Llama, launched earlier this 12 months, containing 65 billion parameters, and OpenAI’s Chat GPT-4, which has not publicly launched its parameters however is rumored to have over 170 trillion.
Parameters matter in terms of LLMs as a result of a rise within the variety of parameters means the mannequin is ready to seize extra particulars within the knowledge. In the end, this helps fashions carry out higher.
Baidu, a distinguished AI developer and web companies supplier in China, launched the newest model of its Ernie chatbot to the general public on Aug. 30. It was among the many first corporations to launch chatbots for public use after the enactment of China’s new AI legal guidelines.
Associated: US and China AI-tech standoff exhibits indicators of spreading to different nations
In keeping with Li, the most recent model of Baidu’s chatbot, Ernie 3.5, is supplied with a processing pace twice that of the earlier model and an improved effectivity of fifty%. He additionally revealed that the corporate plans to launch a more recent model within the “close to future.”
Earlier this summer season, on Aug. 3, the Chinese language e-commerce and tech large Alibaba launched two new AI fashions which might be open-sourced and stated to rival Meta’s Llama 2.
The Chinese language authorities solely not too long ago handed and applied its AI legal guidelines. Beforehand, corporations have been capable of launch small-scale checks of their AI merchandise, which, below the brand new guidelines, widened the check dimension and enabled extra options.
