Baidu Inc (9888.HK), the leading search engine provider in China, has announced that its latest version of the Ernie AI model, called Ernie 3.5, has outperformed the Microsoft-backed OpenAI chatbot and the GPT-4 model in various important metrics.
In a statement on Tuesday, Baidu stated that Ernie 3.5 has achieved higher comprehensive ability scores compared to ChatGPT and demonstrated superior performance in several Chinese language capabilities compared to GPT-4.
Read more | Snapchat AI Threats to Young users
Performance of AI Models
The Beijing-based company referred to a test conducted by the state newspaper China Science Daily, which utilized datasets including AGIEval and C-Eval, two benchmarks commonly used to assess the performance of AI models.
OpenAI has not responded to an email requesting comment during non-business hours.
Read more | Police robots at Singapore Airports
Baidu’s announcement comes at a time when there is a global fascination with AI, sparked by the success of ChatGPT, which has also gained popularity in China. This has led to a surge of domestic companies introducing competing products in the AI market.