How China is building a parallel universe of generative AI

The gigantic leap in technology that machine learning models have shown over the past few months is causing everyone to be excited about the future of AI, but also concerned about its uncomfortable consequences. After text-to-image tools from Stability AI and OpenAI became the talk of the town, ChatGPT's ability to hold intelligent conversations is the new obsession across industries.

In China, where the tech community has always closely watched progress in the West, entrepreneurs, researchers, and investors are looking for ways to enter the generative AI space. Tech companies design tools based on open source templates to appeal to consumers and businesses. Individuals benefit from AI-generated content. Regulators have moved quickly to define how text, image and video synthesis should be used. Meanwhile, US tech sanctions are raising concerns about China's ability to keep up with advances in AI.

As generative AI takes the world by storm towards the end of 2022, let's take a look at how this explosive technology is shaking up China.

Thanks to viral art platforms like Stable Diffusion and DALL-E 2, generative AI is suddenly on everyone's lips. On the other side of the world, Chinese tech giants have also captivated audiences with their equivalent products, adding flair according to the country's tastes and political climate.

Baidu, which has made a name for itself in search engines and has stepped up its game in autonomous driving in recent years, operates ERNIE-ViLG, a 10 billion-parameter model trained on a dataset of 145 million Chinese image-text pairs. How does he behave against his American counterpart? Below are the results of the "kids eating shumai in New York's Chinatown" prompt given to Stable Diffusion, compared to the same prompt in Chinese (纽约唐人街小孩吃烧卖) for ERNIE-ViLG.< /p>

Stable streaming

ERNIE-ViLG

As someone who grew up eating dim sum in China and Chinatowns, I would say the results are on par. Neither got the right shumai, which in the context of dim sum is a type of succulent dumpling, shrimp, and pork in a half-opened yellow wrapper. While Stable Diffusion nails the atmosphere of a Chinatown dim sum restaurant, its shumai is off (but I see where the machine is going). And although ERNIE-ViLG generates a type of shumai, it is a variety more commonly seen in eastern China than the Cantonese version.

The rapid test reflects the difficulty of capturing cultural nuances when the datasets used are inherently biased - assuming that Stable Diffusion would have more data on the Chinese diaspora and ERNIE-ViLG is likely trained on a wider variety of 'shumai images which are rarer outside of China.

Another Chinese tool that has made noise is that of Tencent