[ad_1]
Five months after ChatGPT sparked an investment frenzy in artificial intelligence, Beijing will rein in China’s chatbots, demonstrating the government’s determination to maintain tight regulatory control over the potentially era-defining technology.
China’s Cyberspace Administration announced draft rules this month for so-called generative artificial intelligence — software systems, like the one behind ChatGPT, that can generate text and images in response to users’ questions and queries.
According to the regulations, companies must adhere to the Chinese Communist Party’s strict censorship rules, while websites and apps must avoid publishing material that offends China’s leaders or repeats banned history. The content of the AI system must reflect “socialist core values” and avoid information that undermines “state strength” or national unity.
Companies must also make sure that chatbots create the right words and images and respect intellectual property, and will be required to register algorithms, the software brains behind chatbots, with regulators.
The rules aren’t final, and regulators can keep tweaking them, but experts say engineers building artificial intelligence services in China already know how to incorporate the edict into their products.
All over the world, governments have marveled at the power of chatbots with AI-generated results ranging from the alarming to the mild. Artificial intelligence has been used to crack college exams and create fake photos of Pope Francis wearing a puffy suit.
ChatGPT, developed by the US company OpenAI, which is backed by about $13 billion from Microsoft, has pushed Silicon Valley to apply its basic technology in new areas like video games and advertising. Venture capital firm Sequoia Capital estimates that AI businesses could eventually generate “trillions of dollars” in economic value.
In China, investors and entrepreneurs are racing to catch up. Shares of China’s artificial intelligence company are rising. Splashy announcements have been made by some of China’s biggest tech companies, including most recently e-commerce giant Alibaba; SenseTime, which makes facial recognition software; and Baidu search engine. At least two start-ups developing Chinese alternatives to OpenAI technology have raised millions of dollars.
ChatGPT is not available in China. But faced with many homegrown alternatives, China is quickly opening a red line for artificial intelligence, before other countries are still thinking about how to manage chatbots.
The rules reflect China’s “move fast and destroy things” approach, said Kendra Schaefer, head of technology policy at Trivium China, a Beijing-based consulting firm.
“Because you don’t have a two-party system where both sides argue, they can just say, ‘Okay, we know we have to do this, and we’ll revise it later,'” he added.
Chatbots are trained in large swaths of the internet, and developers are grappling with inaccuracies and surprises from what they sometimes spit out. On the face of it, China’s rules require a level of technical control over chatbots that Chinese tech companies have not yet achieved. Even companies like Microsoft are still perfecting their chatbots to eliminate harmful responses. China has a higher bar, so some chatbots are closed and others are only available to a limited number of users.
Experts are divided on how difficult it will be to train AI systems to be factually consistent. Some doubt that companies can enforce China’s censorship rules, which are often sweeping, ever-changing and even require the censorship of certain words and dates like June 4, 1989, the day of the Tiananmen Square massacre. Others believe that over time, and with enough work, machines can be aligned with certain truths and value systems, even politics.
Analysts expect the rules to undergo changes after consultations with Chinese technology companies. Regulators can ease enforcement so the rules don’t hurt the technology’s development.
China has a long history of internet censorship. Throughout the 2000s, this country has built the world’s most powerful information network through the web. It scares away noncompliant Western companies like Google and Facebook. It hires millions of workers to monitor internet activity.
Meanwhile, Chinese tech companies, which must comply with the rules, are booming, defying Western critics who predict that political control will stifle growth and innovation. When technologies such as facial recognition and mobile phones emerged, companies helped the state to create a surveillance state.
The current wave of AI poses new risks to the Communist Party, said Matt Sheehan, a Chinese AI expert and fellow at the Carnegie Endowment for International Peace.
The unpredictability of chatbots, which will make statements that are nonsensical or false – what AI researchers call hallucinations – runs counter to the party’s obsession with managing what is said online, Mr. Sheehan said.
“Generative artificial intelligence creates a tension between two of the main goals of the party: control of information and leadership in artificial intelligence,” he said.
China’s new regulations are not all about politics, experts say. For example, it aims to protect privacy and intellectual property for individuals and data creators trained by AI models, a topic of worldwide concern.
In February, Getty Images, an image database company, sued artificial intelligence startup Stable Diffusion for training a system to reproduce images on 12 million watermarked photos, which Getty claimed diluted the value of the images.
China is making more of a push to address legal questions about the use of data and content underlying AI companies. In March, as part of a major institutional overhaul, Beijing established the National Data Bureau, in an effort to better define what it means to own, buy and sell data. The state agency will also help the company create the data sets it needs to train the model.
“They now decide what kind of property data and who has the right to use and control,” said Ms. Schaefer, who has written extensively on China’s AI regulations and called the initiative “transformative.”
However, China’s new guardrails may not be timely. The country faces intensified competition and sanctions against semiconductors that threaten to undermine its competitiveness in technology, including artificial intelligence.
Hopes for Chinese AI were high in early February when Xu Liang, an AI engineer and entrepreneur, released one of China’s earliest answers to ChatGPT as a mobile app. The app, ChatYuan, got more than 10,000 downloads in the first hour, Mr. Xu said.
Media reports of marked differences between party lines and ChatYuan’s response immediately surfaced. The response offered a gloomy diagnosis of China’s economy and described Russia’s war in Ukraine as a “war of aggression,” which was inconsistent with the party’s more pro-Russian stance. After a few days, the authorities closed the application.
Mr Xu said he added steps to make the bot more “patriotic”. This includes filtering out sensitive keywords and hiring manual reviewers who can help provide problematic answers. They even trained a separate model that could detect “false viewpoints,” which would be filtered out.
However, it is unclear when Mr. Xu’s bot will satisfy the authorities. The application was originally set to resume on February 13, according to the image, but as of Friday it was still down.
“Services will resume after troubleshooting is completed,” he said.
[ad_2]
Source link