Baidu announced its financial report for the third quarter of 2024 yesterday, showing total revenue of 33.6 billion RMB, a 3% decrease compared to the same period last year. In terms of net profit, Baidu's net profit attributable to the company was 7.6 billion RMB, while the non-GAAP net profit was 5.9 billion RMB. During the subsequent earnings conference call, Baidu's senior management, including Chairman and CEO Li Yanhong, President of the Mobile Ecosystem Group Luo Rong, President of the Intelligent Cloud Group Shen Dou, and Acting CFO He Junjie, provided detailed insights on the highlights of the report.

Recently, Boston Dynamics' robot dog Spot received a significant software update that greatly enhances its ability to navigate complex environments. Although Spot can 'see' the world around it through a 3D vision system, it previously struggled to comprehend obstacles in its surroundings. To address this, the development team introduced a foundation model of artificial intelligence, enabling Spot to recognize and avoid various hazards. With this update, Spot can effectively identify wires, ladders, and rolling objects.

Beijing Zhipu Huazhang Technology Co., Ltd. announced a series of significant technological updates on August 29, 2024, including the release of a new generation foundation model and new application services. At the KDD2024 conference, Zhipu introduced new foundation models including the language model GLM-4-Plus, the text-to-image model CogView-3-Plus, the image/video understanding model GLM-4V-Plus, and the video generation model CogVideoX. These models have achieved international leading standards in their respective fields.

Recently, Microsoft unveiled its latest vision foundation model, Florence-2, which has achieved a significant breakthrough. Leveraging Transformers.js technology, this model can now run 100% locally in browsers that support WebGPU. This breakthrough brings revolutionary changes to AI visual applications, enabling powerful visual recognition capabilities to be realized directly in the user's browser without relying on remote servers.

The nonprofit research organization Machine Intelligence Research Institute (MIRI) is urging a global stop to research on foundation or 'cutting-edge' models, citing concerns that safety issues could threaten human survival. Foundation models are AI systems capable of being applied across various modalities. MIRI believes that foundation models could surpass human intelligence and potentially 'destroy humanity.'

["ZhiYuan Research Institute has released the new generation multimodal foundation model Emu2, pushing the boundaries of multimodal contextual learning capabilities.", "Emu2 surpasses Flamingo-80B and IDEFICS-80B, demonstrating excellent performance in few-shot multimodal understanding tasks.", "Emu2 achieves optimal performance in multiple few-shot understanding, visual question answering, and image generation tasks.", "Emu2-Chat realizes accurate understanding of text-image instructions, while Emu2-Gen offers flexible, controllable, high-quality images."]

IBM and NASA are collaborating to develop geospatial AI to combat climate change. The partnership aims to create multimodal foundation models for weather prediction, heat island effect analysis, mapping, and predicting extreme weather events. IBM plans to open-source the model to promote research and scientific discoveries globally. This foundation model is expected to accelerate scientific discoveries and provide researchers with rapid solutions to contribute to a safer and healthier planet.

Tencent has expanded the parameters of its advertising artificial intelligence model, enhancing targeting and attribution accuracy. Tencent upgraded its self-developed foundation model Tencent Hunyuan and deployed it to internal products. Tencent Cloud is applying the large model across multiple products by upgrading its MaaS capabilities. Tencent Hunyuan plays a role in advertising, gaming, and content sectors. Tencent Cloud promotes the application of large models across various industries.

Zhipu AI unveiled its third generation foundation model ChatGLM3 at the 2023 China Computer Conference. ChatGLM3 features new capabilities including multimodal understanding, code generation and execution, and enhanced web search. After deep optimization, ChatGLM3 has excelled in tests on Chinese and English public datasets. Zhipu Qingyan becomes the first large model product in China with code interaction capability. Zhipu AI looks forward to domestic native large models and domestic chips standing out on the international stage.

["Meta releases open-source version of the computer vision model DINOv2 under the Apache 2.0 license.","DINOv2 can be used for various computer vision tasks and is a flexible and versatile tool.","Meta plans to combine DINOv2 with large-scale language models to create more powerful image analysis systems."]