Taipei, Taiwan – As chatbots powered by artificial intelligence send shockwaves through the global tech industry, China is racing to produce versions of its own.
China’s search engine giant Baidu has announced plans to release its own chatbot, ERNIE, sometime in March, following the successful launch of ChatGPT, which has sparked existential questions about the future of sectors from education to journalism and health care.
Similar projects are underway at tech giants Huawei, Alibaba, Tencent, JD.com and leading institutions including the Beijing Academy of Artificial Intelligence.
Chinese technology shares have rallied in response to the news and authorities have pledged to step up support for the sector. China’s Ministry of Science and Technology said last week it would push for the integration of AI across Chinese industry, while cities including Beijing have also announced plans to support – development.
But while China looks set to produce a fast successor to compete with ChatGPT, developed by California-based OpenAI, there are big questions about how it will work the technology within an ecosystem that includes strict controls on the Internet.
“The most general purpose technology we have, artificial intelligence, should be very general purpose,” Jeffrey Ding, an assistant professor at George Washington University who studies China’s tech sector, told Al Jazeera.
“But it is really shaped by the specific, political, cultural, linguistic context in which these models are developed and used.”
Bots like ChatGPT rely on generative AI to generate responses that draw from billions of data points scraped from the internet, which also makes their responses at times difficult to predict.
Long conversations between ChatGPT and users have gone off the rails, forcing Microsoft to limit its ChatGPT-powered Bing search engine to a maximum of five queries to keep it working. ChatGPT’s responses have also rankled U.S. conservatives, who have accused the bot of “waking up” on hot-button social issues like affirmative action and transgender rights.
In China, internet censors routinely block keywords, delete posts, and ban users based on the sensitivities of the Chinese Communist Party (CCP), forcing creative users between -use web homophones, coded messages and screenshots to get around information controls.
For a chatbot, the censorship tool means a very limited collection of information to rely on.
Baidu’s ERNIE chatbot is based on information scraped from inside and outside the Chinese firewall – necessary to obtain a relevant data set – and draws on sources such as Wikipedia and the Notoriously unwise reddit.
Assuming their products are technologically capable of performing at the same level as ChatGPT, Chinese tech companies may find themselves choosing between limiting what chatbots, such as Bing, can do Microsoft, or what they may say.
“It will make it much more useful, but it will make it a little bit safer politically,” Matt Sheehan, another fellow at the Carnegie Endowment for International Peace who studies AI and China, told Al Jazeera.
“Historically, almost every time they’ve faced a trade-off between information control and … business opportunities, they always come down on the information control side, and then they ‘assume businesses will figure it out.’
In 2017, Tencent pulled two chatbots from its QQ messaging app after they allegedly made comments that were considered unpopular. One chatbot, developed by Microsoft, told users it dreamed of moving to the US, while the other chatbot developed by Chinese tech company Turing Robot told users He used to dislike the CCP.
More recently, YuanYu Intelligence, a startup based in Hangzhou, suspended its chatbot earlier this month after it gave negative answers about the Chinese economy, although the company’s lead developer in interview with the Washington Post that the ban was simply due to technical errors. .
Baidu itself has previously anticipated Beijing’s red lines, as shown by the ERNIE VilG Image and art generator, which was released in demo form last year.
Although widely praised for performing as well or better than Western competitors, the app blocks users from content related to politically sensitive topics such as Tiananmen Square, democracy, Xi Jinping and Mao Zedong.
“With generative AI, the power of the machine is its ability to be creative and connect things that you wouldn’t expect to be connected, and do things in different styles that are expected,” Sheehan said.
“But how can you prevent criticism that may be more subtle or less direct about the core beliefs of the Communist Party without completely destroying the machine itself? That seems like a very difficult technical and socio-political problem.”
Before the release of ChatGPT, China was already taking steps to regulate AI. On Wednesday, the Cybersecurity Administration began implementing new rules governing search engine recommendations, giving users more control over how their personal data is used by search engines.
In January, China also passed legislation to regulate deep synthesis – a type of generative AI that can be used to create “deep fakes” – and last year established a record for algorithms, although the impact is long -expected time of both steps broadly. seen vaguely.
As part of a wider crackdown on the tech industry from 2020, authorities pulled the plug on major IPOs by Ant Group, financial affiliate Alibaba, and ride-hailing app Didi over alleged data concerns.
Despite being blocked by the Chinese firewall, ChatGPT has created a huge buzz among Chinese users accessing the site through virtual private networks (VPNs) and other roundabout methods .
Much of that buzz has come from ChatGPT’s ability to perform in Chinese and other languages despite being trained in English, said Ding, the George Washington University professor.
“The excitement is not just about business applications. Part of it is just excitement and wonder at how amazing the natural language capabilities of this technology are,” he said.
“And one aspect of that is that ChatGPT was not even trained on any Chinese language texts. Most of it was trained on English texts, but I’ve seen Chinese users ask questions in Mandarin, and it still performs very well in another language.”
Still, Chinese could be particularly challenging for AI, Ding said, because of the language’s heavy use of idioms and expressions with historical context.
While Chinese developers have already released several chatbots, including Inspur’s Yuan 1.0 and Fudan University’s MOSS, neither has come close to matching ChatGPT’s capabilities.
Unlike Silicon Valley, Chinese tech companies have so far focused on consumer-facing products with a short development cycle, said Chim Lee, a China tech analyst at the Economist Intelligence Unit, putting them at a disadvantage. in an older field like AI.
The arrival of ChatGPT provided Chinese companies with “proof of concept,” Lee said, showing both the promise of next-generation AI and the need for long-term investment.
“Baidu has been considering this model for a long time, but you have to justify this kind of investment just to train the model, let alone analyze or discuss the long-term foundational data.” time associated with the algorithm,” said Lee. Al Jazeera.
“What’s really helpful with ChatGPT now is that these companies can say, ‘Hey, we want to develop these things and they can tell the government that’s what I want to do. do.'”
Rui Ma, tech analyst and creator of Tech Buzz China, said it’s anyone’s guess which Chinese company could come out on top in the race to match ChatGPT, although Baidu appears to be a -out of the gate first.
“I think right now the biggest excitement is still at the model level,” Ma said.
Alibaba told Al Jazeera that it is testing the GPT Chat-style bot internally for use in its apps and cloud services, but did not provide further details or respond. e questions about censorship.
JD.com led Al Jazeera to a statement released last week about its plans to deploy its ChatJD business chatbot for use on its retail and financial website, based on 10 years of data from its various platforms.
Baidu, Tencent and Huawei did not respond to requests for comment.
In addition to Beijing’s scrutiny, Chinese tech companies also face obstacles from abroad in the form of export controls.
In August, US President Joe Biden signed the CHIPS and Sciences Act, which requires tech companies that receive government subsidies to move the manufacturing of advanced chips out of China.
Although Chinese tech companies have strategic sources of chips, Washington’s attempt to corner the sector poses a long-term threat, said the EIU’s Lee.
“The US specifically banned the export of these very advanced AI chips that would be used in model training, or even just employment, so all these factors put Chinese AI developers in an unfair situation in many ways,” he said.
“Many Chinese companies and research centers have actually assembled some chips that would be used for this kind of applications, but if you look at the amount of chips that ChatGPT needs, there is a big possibility that it will not run the chips out at some. point,” he said.
