The diverging path of China's two leading AI players shows where the country's artificial intelligence industry is headed. DeepSeek and ByteDance, the two leaders of China's AI industry, are adopting vastly different strategies. On Monday, DeepSeek released DeepSeek V3.2, another open-weight model that anyone can tinker with. The startup says it performs on par with the latest models from OpenAI and Google, and it even beats them on some key mathematics benchmarks. That same day, ByteDance, whose dominance in AI applications we covered previously, introduced ways for people to use its chatbot, Doubao.

ByteDance's Other AI Chatbot Is Quietly Gaining Traction Around the World ByteDance is paying for ads and partnering with influencers to promote its AI chatbot app Cici in countries like the UK, Mexico, and Indonesia. ByteDance, the parent company of TikTok, has built what is currently the most popular AI chatbot in China: Doubao . Launched in 2023, the app has risen to the top of the country's generative AI market, reaching more than 157 million monthly active users by August, according to Chinese analytics firm QuestMobile. But what's less known is that Doubao also has an overseas counterpart: Cici. It was released around the same time and features a nearly identical female cartoon avatar as its app icon, except Cici's has longer hair than Doubao's.

ByteDance's Doubao app has overtaken DeepSeek, proving that user-friendly design often matters more than having the most advanced AI model. When Chinese AI startup DeepSeek became a global sensation in January, it not only shocked Silicon Valley but also startled ByteDance, TikTok's parent company. The Chinese tech giant had already launched Doubao, its own flagship AI assistant app with tens of millions of users. But when DeepSeek became the best-known Chinese AI company overnight, no one was talking about Doubao anymore. Now, ByteDance has gotten its revenge.

In 2025, Large Language Model (LLM) services have launched a new feature -- AI video chat -- allowing users to interact with AI agents via real-time video communication (RTC), just like chatting with real people. Despite its significance, no systematic study has characterized the performance of existing AI video chat systems. To address this gap, this paper proposes a comprehensive benchmark with carefully designed metrics across four dimensions: quality, latency, internal mechanisms, and system overhead. Using custom testbeds, we further evaluate five mainstream AI video chatbots with this benchmark. This work provides the research community a baseline of real-world performance and identifies unique system bottlenecks. In the meantime, our benchmarking results also open up several research questions for future optimizations of AI video chatbots.

With the rapid development of LLM-based agents, increasing attention has been given to their social interaction and strategic reasoning capabilities. However, existing Werewolf-based benchmarking platforms suffer from overly simplified game settings, incomplete evaluation metrics, and poor scalability. To address these limitations, we propose WereWolf-Plus, a multi-model, multi-dimensional, and multi-method benchmarking platform for evaluating multi-agent strategic reasoning in the Werewolf game. The platform offers strong extensibility, supporting customizable configurations for roles such as Seer, Witch, Hunter, Guard, and Sheriff, along with flexible model assignment and reasoning enhancement strategies for different roles. In addition, we introduce a comprehensive set of quantitative evaluation metrics for all special roles, werewolves, and the sheriff, and enrich the assessment dimensions for agent reasoning ability, cooperation capacity, and social influence. WereWolf-Plus provides a more flexible and reliable environment for advancing research on inference and strategic interaction within multi-agent communities. Our code is open sourced at https://github.com/MinstrelsyXia/WereWolfPlus.

Large Language Models (LLMs) are increasingly used as automated evaluators in natural language generation, yet it remains unclear whether they can accurately replicate human judgments of error severity. In this study, we systematically compare human and LLM assessments of image descriptions containing controlled semantic errors. We extend the experimental framework of van Miltenburg et al. (2020) to both unimodal (text-only) and multimodal (text + image) settings, evaluating four error types: age, gender, clothing type, and clothing colour. Our findings reveal that humans assign varying levels of severity to different error types, with visual context significantly amplifying perceived severity for colour and type errors. Notably, most LLMs assign low scores to gender errors but disproportionately high scores to colour errors, unlike humans, who judge both as highly severe but for different reasons. This suggests that these models may have internalised social norms influencing gender judgments but lack the perceptual grounding to emulate human sensitivity to colour, which is shaped by distinct neural mechanisms. Only one of the evaluated LLMs, Doubao, replicates the human-like ranking of error severity, but it fails to distinguish between error types as clearly as humans. Surprisingly, DeepSeek-V3, a unimodal LLM, achieves the highest alignment with human judgments across both unimodal and multimodal conditions, outperforming even state-of-the-art multimodal models.

Artificial Intelligence (AI) writers have emerged as a signi ficant force in the realm of content creation. These advanced tools leverage natural language processing techniques to g enerate coherent and logical texts, applicable across vari ous domains such as journalism, advertising, and educational m aterials. This document delves into the capabilities, applications, and implications of AI writers, examining thei r technological underpinnings, market influence, strength s, limitations, future trajectories, and ethical considerat ions. In the rapidly evolving landscape of artificial intelligenc e technologies today, AI models are increasingly being appl ied across various domains, with literary creation being no exc eption.

Baidu, which is often called China's answer to Google, has launched its own pair of AI-powered smart glasses at its annual World Conference event in Shanghai. The device will run on the company's ERNIE generative AI technology and was designed to "become a private assistant," according to the Financial Times. Users will reportedly be able to interact with the device using their voice and ask it questions about what it sees in their current environment. They can also tell it to play music and even track their calories consumption. And since the glasses are equipped with cameras, they can ask it to snap photos or take videos.

ByteDance is pushing further into hardware with its first set of earbuds. The TikTok owner's 170 Ola Friend buds will only be available in China for now. They link to the company's generative AI assistant, Doubao, which can be activated with a voice command. According to the South China Morning Post, wearers will be able to use the open-ear buds as a tour guide while traveling and to practice English (among other things) after they're paired with the Doubao mobile app. The publication notes that Doubao is currently the most popular GAI app in China, with more than 47 million monthly active users in September.