OpenAI’s latest blunder shows the challenges facing Chinese AI models

The truth is, among the many few lengthy Chinese language tokens in GPT-4o that aren’t both pornography or playing nonsense, two are “socialism with Chinese language traits” and “Folks’s Republic of China.” The presence of those phrases suggests {that a} important a part of the coaching knowledge really is from Chinese language state media writings, the place formal, lengthy expressions are extraordinarily frequent.

OpenAI has traditionally been very tight-lipped concerning the knowledge it makes use of to coach its fashions, and it in all probability won’t ever inform us how a lot of its Chinese language coaching database is state media and the way a lot is spam. (OpenAI didn’t reply to MIT Know-how Overview’s detailed questions despatched on Friday.)

However it’s not the one firm battling this drawback. Folks inside China who work in its AI business agree there’s a scarcity of high quality Chinese language textual content knowledge units for coaching LLMs. One motive is that the Chinese language web was once, and largely stays, divided up by massive firms like Tencent and ByteDance. They personal many of the social platforms and aren’t going to share their knowledge with rivals or third events to coach LLMs. 

The truth is, that is additionally why serps, together with Google, kinda suck in relation to looking in Chinese language. Since WeChat content material can solely be searched on WeChat, and content material on Douyin (the Chinese language TikTok) can solely be searched on Douyin, this knowledge just isn’t accessible to a third-party search engine, not to mention an LLM. However these are the platforms the place precise human conversations are occurring, as a substitute of some spam web site that retains attempting to attract you into on-line playing.

The shortage of high quality coaching knowledge is a a lot larger drawback than the failure to filter out the porn and normal nonsense in GPT-4o’s token-training knowledge. If there isn’t an present knowledge set, AI firms must put in important work to determine, supply, and curate their very own knowledge units and filter out inappropriate or biased content material. 

It doesn’t appear OpenAI did that, which in equity makes some sense, given that folks in China can’t use its AI fashions anyway. 

Nonetheless, there are a lot of folks dwelling outdoors China who need to use AI companies in Chinese language. And so they deserve a product that works correctly as a lot as audio system of some other language do. 

How can we resolve the issue of the shortage of fine Chinese language LLM coaching knowledge? Inform me your concept at zeyi@technologyreview.com.