Baidu made an exciting announcement on June 27th, expressing its delight in the superior performance of its latest AI foundation model, Ernie 3.5, compared to ChatGPT in certain key areas.
According to a test conducted by China Science Daily, a state-run newspaper, Ernie 3.5 showcased better overall capabilities than ChatGPT based on GPT-3.5, highlighting its advancements.
Similar News… NYC Lawyers Fined $5000 for ChatGPT-Generated Briefs
Baidu further emphasized that Ernie 3.5 outperformed the more advanced GPT-4 in various Chinese language capabilities, as stated in their official statement.
To evaluate the performance of AI models, Baidu employed datasets such as AGIEval and C-Eval, which served as benchmarks during the testing process.
Baidu enthusiastically stated, “Ernie 3.5 has made significant improvements in effectiveness, functionality, and performance.”
Comparing it to its previous version, Ernie 3.0, the company emphasized that the new model offers a remarkable two-fold increase in training throughput and an astonishing 17-fold increase in inference throughput.
Introduced in February and released in March as a competitor to OpenAI’s ChatGPT and Google’s Bard, Ernie Bot has been undergoing public beta testing. It’s worth noting that both ChatGPT and Bard are inaccessible in China.
Following Baidu’s lead, other prominent Chinese tech giants, including Alibaba and Tencent, have also launched their own AI products.
As a leading internet company in China, Baidu has made significant investments in AI and is recognized as the creator of PaddlePaddle, the country’s first open-source deep-learning platform.
Earlier this year, Baidu unveiled its plans to carve a niche in the saturated smartphone market by introducing a device integrated with AI features.
Moreover, Baidu is seamlessly integrating Ernie Bot into its broader ecosystem of devices, such as smart speakers.