Tech On Me: Are Chinese tech giants doing enough to eliminate xenophobia?

This week's focus 

China's tech giants, including Tencent and NetEase, have announced efforts to curb ‘extreme nationalism’ online, particularly anti-Japanese sentiment, following the fatal stabbing of a Chinese woman who protected a Japanese mother and child in Suzhou. 

Tencent-owned WeChat and NetEase stated they would investigate and ban users inciting hatred. 

The incident in Jiangsu province drew significant public attention, with some netizens provoking anti-Japanese sentiments. The attacker, an unemployed man named Zhou, stabbed the Japanese mother and child outside a Japanese school, and the Chinese woman who intervened, Hu Youping, succumbed to her injuries.

My take: There should be no place for hate and discrimination on social media, so I am glad to see Chinese tech giants take action. In recent years, nationalist sentiment against Japan has been prevalent on China's Internet, with little intervention. 

Unfortunately, social media content mocking Japanese schools is very common, and many believe such xenophobic content fuels real-life attacks. After the Suzhou incident, Weibo, a platform with 588 million monthly active users, removed 759 pieces of illegal content, and Tencent addressed 836 posts violating its rules. Both platforms blocked several accounts. 

However, it is disappointing that while Hu was praised for her bravery, there was also an extreme nationalist reaction against her. Weibo noted some users posted remarks inciting nationalist sentiment and group hatred. Douyin, the Chinese version of TikTok, mentioned it would look into ‘extreme xenophobia’ on specific accounts.

In other news

No social media for Australian teens?

The Australian government has extended its trial of age assurance technology, initially aimed at blocking children from accessing pornographic content, to now also consider restricting younger teenagers from social media platforms. The $6.5 million trial, announced in the federal budget, will now evaluate how accurately different technologies can determine ages below 18.

The communications department is consulting with experts, children, and parents to gather concerns and interests, aiming to provide policy advice to the government. Meta, the owner of Facebook and Instagram, has shown reluctance to verify user ages, suggesting that age assurance should be managed by app stores like Apple and Google.

The department engages with social media companies, although they are not legally obliged to participate. The trial's findings will inform the enforcement of new online safety codes. The trial will also assess the effectiveness of various technologies, including their ability to be circumvented via virtual private networks (VPNs).

My take: The government's move to protect children from exposure to inappropriate content and potentially harmful interactions online is commendable. This is especially relevant given the rising concerns about mental health and online safety among younger users. 

However, implementing age verification, primarily through facial recognition or other biometric methods, raises significant privacy concerns. Data breaches and misuse of minors' personal information are risks which need careful consideration and stringent safeguards. I am also doubtful about the effectiveness of these technologies, especially their accuracy in determining the ages of younger teens, as users may circumvent age restrictions through VPNs or other means.

Ultimately, with the lack of legal obligation, I do not see active industry participation and compliance as the likes of Meta will do all it takes to resist.

Should Meta charge for ad-free social networks?

The European Commission has accused Meta, the parent company of Facebook and Instagram, of violating the EU’s Digital Markets Act (DMA) with its 'pay or consent' advertising model. Meta launched this model in 2023 and offered users a choice between paying a monthly fee for an ad-free experience or consenting to use their data for targeted ads.

The European Commission's preliminary investigation found that this model does not comply with the DMA. The commission argues that the binary choice forces users to consent to extensive data collection across multiple platforms if they choose not to pay. Additionally, Meta does not offer a less data-intensive option comparable to the ad-supported versions of Facebook and Instagram.

According to the commission, Meta must provide ‘equivalent’ versions of its social networks that use less personal data to comply with the DMA. A spokesperson for Meta stated that the subscription model was designed to meet various regulatory requirements, including those of the DMA, and expressed the company's intention to continue engaging constructively with the commission.

My take: Meta's ‘pay or consent’ model attempts to comply with regulatory demands by offering a choice to users. However, the commission has made an essential point that this model forces users into a binary option, effectively coercing them to consent to extensive data collection if they don't pay. 

As a consumer and a journalist covering the media and tech industry, I understand businesses need the freedom to innovate and grow. However, this growth must not come at the expense of user rights and privacy. As I am not a fan of platforms self-regulating, regulatory bodies play a vital role in maintaining the balance of user rights and privacy, ensuring that technological advancements and business models align with ethical standards and consumer protection.

Baidu’s Ernie Bot reaches 300 million users

China’s Baidu has announced that its ChatGPT alternative, Ernie Bot, now has 300 million users. This milestone was revealed at Wave Summit 2024, where Baidu introduced the latest version of its AI foundation model, Ernie 4.0 Turbo. Ernie Bot reached 100 million users by December 2023 and doubled that number by April 2024. The bot handles 500 million queries daily. The new model is available via Baidu’s official website and app, and corporate clients can access it through the AI Cloud Qianfan platform.

My take: With such a large user base, data privacy and security will become paramount. I want to see how Baidu manages and protects user data, which is crucial, especially given China's stringent data regulations and the global concerns around data security in AI applications. While handling 500 million queries a day is a significant technical feat, it will be necessary for Baidu to ensure the AI provides accurate, unbiased, and helpful responses.

Zhihu introduces generative AI Q&A tool 

Zhihu has introduced a new AI-driven tool named Zhida.ai, which provides users with reliable information derived from quality content within the Zhihu community. This tool by China's equivalent of Quora, is developed in partnership with domestic AI startup Model Best, utilising the Zhihaitu AI foundation model launched in April. When users pose questions to Zhida.ai, the tool responds with a summary, a detailed answer, and a list of sources, including direct links to relevant Zhihu pages.

My take: Aside from the interface looking much like ChatGPT, Zhida.ai is part of a broader trend among Chinese tech companies like Baidu (as mentioned above), Tencent and Alibaba. Tencent has integrated its Hunyuan LLM into over 180 services, including Tencent Meeting and WeChat, while Alibaba's Tongyi Qianwen LLM is used by over 90,000 corporate clients via the DingTalk app. It is an exciting time for generative AI innovation in China.

Generative AI in metaverse games?

Meta is planning to integrate more generative AI technology into its VR, AR, and mixed-reality games as part of its effort to revitalise its metaverse strategy. The company seeks to explore "new consumer experiences" through generative AI, aiming to create games that offer unique experiences each time they are played. Meta plans to develop or partner with third-party creators to build AI-powered tools to improve game development efficiency and speed.

The focus will be on Horizon, Meta’s suite of metaverse games and apps, but the initiative might extend to other platforms like smartphones and PCs. This new approach has the potential to significantly enhance the gaming ecosystem by increasing efficiency and content creation capabilities.

My take: This move by Meta makes sense to me as the platform has already invested heavily in its metaverse projects with limited financial returns. Meta’s Reality Labs, responsible for its metaverse projects and the Quest headset, faces challenges in attracting users to its Horizon platform and recovering from significant financial losses. Generative AI is increasingly used in game development, with various companies employing the technology to enhance game dialogues and narratives, so Meta has made a good move by allowing third-party manufacturers to use some of Quest’s software features. 

Looking ahead

Game Changers and Tech MVP 

In my blog last week, I announced that Tech MVP will return for another year. I have some news to share—Tech MVP winners will be announced at Campaign’s Game Changers 2024. This means more well-deserved recognition for the winners. I am excited!