Domestic artificial intelligence company Wenwen Xinqun has announced the open sourcing of its latest developed end-side full-modal understanding AI model — Megrez-3B-Omni. This model is the world's first open source project of its kind, marking the company's innovative development in the AI field. At the same time, Wenwen Xinqun also launched a pure language version of the model, Megrez-3B-Instruct, to further enrich its product line. Founded in May 2023, the founding team comes from the Department of Electronic Engineering at Tsinghua University. The company is committed to creating efficiency.

Shanghai Leap Star Intelligent Technology Co., Ltd. recently announced that its smart life assistant app 'Leap Inquiry' has implemented a one-click access feature for the smart visual search function 'Photo Inquiry' on the iPhone 16. Users can now easily set up and directly activate 'Photo Inquiry' using the 'Camera Control' button located at the bottom right side of the iPhone 16. This innovative attempt aims to provide a smoother user experience. Additionally, users of other Apple models can also quickly access this feature through the control center and lock screen shortcuts after upgrading to the latest iOS 18 version.

The rapid development of generative AI has raised the challenge of comprehensively evaluating its performance. A plethora of models have emerged, each showcasing increasingly impressive results. However, the pressing question remains: how do we assess the effectiveness of these text-to-image models? Traditional evaluation methods often rely on subjective visual assessments or simple metrics like CLIPScore, which frequently fail to capture the intricate details found in complex text prompts, such as relationships between objects and logical reasoning. This leads to inaccurate evaluation outcomes for many text-to-image models.

With the development of AI technology, many difficult problems in learning scenarios are being solved efficiently. On November 1, Quark fully upgraded its 'AI Question Search' product, making the process of searching and solving problems faster and more capable. Users can find various new and difficult questions through their smartphones, tablets, and computers, and the AI will provide step-by-step explanations, allowing users to ask questions at any time. After launching the 'One-Stop AI Services' and 'System-Level All-Scenario AI Capabilities', Quark's AI capabilities are being applied to learning scenarios, helping users enhance their learning abilities and efficiency.

Recently, Ali Tongyi Qianwen officially responded to rumors regarding its advertising expenditure. Previously, there were online reports claiming that Tongyi Qianwen's advertising costs for the third quarter reached as high as 4.5 million, with 900,000 spent in the first 20 days of October. In response, Tongyi Qianwen issued a statement on Weibo denying these figures, stating that the actual advertising expenditure is far lower than what has been rumored.

Recently, a seemingly trivial question has sparked heated discussions worldwide: Are there more wheels or doors in the world? While this question may appear inconsequential at first glance, it actually provides an excellent opportunity to explore the capabilities of generative artificial intelligence (AI). This topic unexpectedly went viral in 2022, with a simple question igniting countless debates on Twitter and subsequently gaining widespread attention on TikTok. People began to fervently discuss which one has a greater quantity: wheels or doors. However, answering this question...

In today's rapidly advancing technology landscape, artificial intelligence (AI) is infiltrating various aspects of our lives at an unprecedented speed. However, the application of this new technology in the field of education has sparked a controversy, and a case in Hingham, Massachusetts is a typical example. The protagonists of the incident are a senior student at Hingham High School. According to his mother, Jennifer Harris, her son used AI to gather research materials and create an outline while working on a group project with a classmate.

Meta's Chief Technology Officer Andrew Bosworth recently confirmed an important decision to The Verge: the company has canceled the high-end headset Quest Pro2 (codename La Jolla), originally scheduled for release in 2027, and is now developing a new ultra-lightweight headset. This news marks a significant change in Meta's hardware strategy and reveals a new direction for the company's VR technology development. According to Bosworth, the main reason for canceling Quest Pro2 was primarily...

Recently, Apple's AI research team launched their next-generation family of Multimodal Large Language Models (MLLMs) - MM1.5. This series of models can integrate various data types such as text and images, showcasing new capabilities of AI in understanding complex tasks. Tasks like visual question answering, image generation, and multimodal data interpretation can be better addressed with the help of these models. A major challenge for multimodal models is how to achieve effective interaction between different data types. Previous models often struggled in this aspect.

Recently, the U.S. Secret Service quietly ventured into the field of artificial intelligence, spending over $50,000 on cloud services from Microsoft Azure and OpenAI. This move has drawn significant attention, but the Secret Service has remained silent, not revealing specific purposes or plans for the usage. This information was disclosed through documents obtained by the media, highlighting the Secret Service's growing interest in AI technology. Image source note: The image was generated by AI, with licensing provided by Midjourney.

Recently, a study published in Nature revealed a thought-provoking phenomenon in the development of artificial intelligence: as large language models (LLMs) continue to evolve, they increasingly tend to provide answers with 'full confidence', even when those answers may be incorrect. This phenomenon has sparked broad discussions on AI reliability and the risks of use. The research team, led by José Hernández-Orallo from the Valencia Institute of Artificial Intelligence in Spain, conducted an in-depth analysis of AI models.

At the Meta Connect conference on Wednesday, Meta marketed its Quest mixed reality platform as the next major app store for developers. Meta has stated that it is working to make it easier for mobile developers to transition to Meta's Horizon OS. Meta Vice President Mark Rabkin announced that if developers can build apps for Android, they can also create applications for Meta Horizon OS using their preferred languages and IDEs.

Recently, a Reddit user shared his experience on the forum using OpenAI's latest model, OpenAI o1, to solve high school math problems. Remarkably, he got all the answers correct. @FitAirline8359 mentioned using past math final exam questions from Chinese universities' upperclassmen and presented these problems to OpenAI's latest version, o1. To his surprise, OpenAI o1 answered all of them correctly in just 20 minutes. This result left him extremely shocked and sparked significant discussion among other users.