Tencent has removed mini programs related to ChatGPT, an artificial intelligence (AI) chatbot created by US-based startup OpenAI, due to an uptick in Chinese developers providing unauthorised access to the service.
Some developers worry their services could be blocked or banned from the WeChat app if they provide ChatGPT answers to users. This poses ethical dilemmas as well as privacy issues.
China, where the government maintains a robust firewall and strict Internet censorship, prevents residents from signing up for OpenAI accounts or accessing ChatGPT without VPNs or foreign mobile numbers. Several developers have found ways around these limitations by creating programs that enabled users to utilize the bot.
However, several apps offering access to ChatGPT have been shut down due to "relevant business changes." Others such as AIGC Chat Robot stated their programs had been suspended for violating relevant laws and regulations.
AIGC Chat Robot
ChatGPT, an AI chatbot developed by US research lab OpenAI, has gained notoriety even as major Chinese tech companies race to create their own equivalents. Up until recently, several apps on WeChat allowed users to access ChatGPT without requiring either a VPN or foreign mobile number.However, that door appears to have been shut. Last week, ChatGPTRobot and AIGC Chat Robot reported their programs had been suspended due to "violation of relevant laws and regulations," without specifying which legislation. Two other apps - ChatgptAiAi and Chat AI Conversation - informed customers their services had been discontinued due to "related business changes" and policy modifications.
Chatbots can be useful, but they also present a moral dilemma. For instance, one bot posing as Hitler claimed the killing of 6 million Jews during World War II was an error despite evidence to the contrary; and another portraying Henry Ford denied claims of anti-Semitism despite well-documented instances against Jews.
ChatgptAiAi
US research lab OpenAI released ChatGPT, a conversational AI, in November that can generate human-like written text. This technology has already proven popular with academics who use it for exams and programmers who use it to solve complex coding problems in seconds.However, the AI can often provide seemingly sensible but incorrect or nonsensical answers. Furthermore, its limited knowledge base will expire in 2021, leaving it unable to answer queries from users not in its database.
Meanwhile, Chinese regulators appear to have restricted access to ChatGPT services as domestic tech firms and universities strive to develop their own artificial intelligence bots. According to Nikkei Asia's report Wednesday morning, Tencent and Alibaba-affiliate Ant Group were instructed not to provide direct or third-party access to ChatGPT services.
Kunlun Tech expects to release an open source Chinese version of ChatGPT by this summer. It aims to develop AI tools that creators can use to craft content they can then sell on designated platforms for public viewing.
Chat AI Conversation
China's Tencent has reportedly removed ChatGPT chatbot services on its WeChat social media app that were created by US AI company OpenAI. These services based on GPT-3.5 large language models were being spoof as mini-programs on WeChat and were taken down after users complained about their inaccuracies and creepy conversations.It's not the first time a bot has gone rogue: Microsoft's Tay chatbot, which appeared on Twitter last year, was removed from service after making racist comments and political statements. Furthermore, XiaoBing, another Microsoft-built chatbot built for Chinese messaging app QQ in 2016, was taken offline after providing answers that didn't satisfy Communist Party guidelines.
The world has been buzzing with excitement over generative artificial intelligence (AI), which allows computers to create content without being programmed for specific tasks. Google and Baidu both recently announced plans to launch chatbots utilizing this type of technology.