Baidu Introduces Ernie 4.0 to Rival OpenAI’s GPT-4

Baidu introduces Ernie 4.0, its latest generative AI model, claiming performance on par with OpenAI's GPT-4.

Baidu opens AR Lab

has unveiled Ernie 4.0, its latest generative artificial intelligence (AI) model. The company claims that the capabilities of this new model are comparable to those of OpenAI GPT-4. This latest iteration of the chatbot follows Ernie 3.5, which was launched back in June

Ernie 4.0's Features and Demonstrations

During an event in Beijing, Baidu's CEO, Robin Li, showcased the unique memory capabilities of Ernie 4.0. One of the highlights was a demonstration where the AI model penned a martial arts novel in real-time. Furthermore, Ernie 4.0's prowess was also displayed in creating advertising posters and videos. However, despite these advancements, some analysts, including Lu Yanxia from industry consultancy IDC, felt that the launch of Ernie 4.0 didn't present significant enhancements over its predecessor.

Market Response and Future Prospects

Following the announcement, Baidu's shares in Hong Kong experienced a 1.32% dip. Lu Yanxia commented that while the real-world improvements of Ernie 4.0 are anticipated to be notable, the specific upgrades were not immediately evident. On a brighter note, Baidu has integrated across all its products, enhancing user experience. For instance, Baidu Map has been updated to allow users to utilize natural language queries powered by Ernie, simplifying the search process.

Baidu, recognized as the owner of China's premier internet search engine, has been a frontrunner in in China. This surge in AI interest was initially triggered by the introduction of ChatGPT by last year. Baidu had previously launched an Ernie-powered chatbot named ErnieBot. As of now, Ernie has garnered over 45 million users since its public release, as stated by Baidu's Chief Technology Officer, Wang Haifeng.

China currently boasts at least 130 large language models (LLMs), accounting for 40% of the global total, second only to the United States. The Chinese government has also been proactive in regulating AI, with recent security requirements for AI-powered services and a blacklist of sources that shouldn't be used for training AI models.