Search AI Products and News

Recently, tech giant NVIDIA, in collaboration with the Massachusetts Institute of Technology (MIT) and the University of Hong Kong, released a new framework called Fast-dLLM. This innovative framework aims to significantly increase the inference speed of diffusion models (Diffusion-based LLMs), up to 27.6 times, providing stronger technical support for AI applications. The challenges and opportunities of diffusion models make them strong competitors to traditional autoregressive models.

According to Legal Daily, the Suzhou Intermediate People's Court in Jiangsu Province recently issued a final ruling in a copyright dispute over AI-generated images. The court ruled that the images of the "Illusion Wings Transparent Art Chair" series, generated by plaintiff designer Feng Moumou using AI software, do not constitute works in the sense of copyright law. The court therefore rejected Feng's lawsuit against the defendants, Zhu Moumou et al., for infringement. Plaintiff Feng Moumou created a series of art chair images using an AI image generation software and published them on social media seeking mass production cooperation. Defendant Zhu Moumou contacted the plaintiff seeking cooperation but was refused. Afterwards...

Recentemente, a Microsoft lançou gratuitamente no seu navegador Edge o novo recurso de assistente de IA, o Copilot Vision. Este recurso oferece aos usuários uma experiência mais inteligente, permitindo a interação por voz com o assistente de IA para interpretar o conteúdo na tela em tempo real, ajudando os usuários a concluir várias tarefas. De acordo com Mustafa Suleyman, CEO de IA da Microsoft, o Copilot Vision é um recurso de assistente baseado em voz. Para usar, os usuários simplesmente precisam ativar o recurso,

Shopify CEO Tobi Lütke issued a memo instructing teams to demonstrate why AI cannot perform the required work before seeking additional personnel or resources. The memo aims to encourage exploration of AI integration into workflows and spur innovation. Image note: Image generated by AI, licensed through Midjourney. Lütke's memo further...

With the application of DeepSeek, Baidu Maps has completed its deep integration with DeepSeek-R1. Users can directly ask questions in the search bar, such as inquiring 'What fun places are there in Zhengzhou?' By selecting the DeepSeek tag, they can receive answers that have been deeply analyzed. These answers not only provide recommended attractions but also include relevant addresses, allowing users to navigate with a click. As a result, the efficiency of obtaining information while using the map has greatly improved. It's worth mentioning that Baidu has become one of the successors to Tencent.

In multimodal tasks, visual language models (VLMs) play a crucial role, such as in image retrieval, image captioning, and medical diagnosis. These models aim to align visual data with language data for more efficient information processing. However, current VLMs still face significant challenges in understanding negation. Negation is critical in many applications, such as distinguishing between "a room without windows" and "a room with windows." Despite significant progress made by VLMs, existing models fall short when it comes to handling negated statements.

Li Xiang, founder and CEO of Ideal Automotive, is set to return to the public eye after a period of silence online. Ideal Automotive officially announced that from December 25 to 27, Li Xiang's dialogue program will be broadcast over three consecutive nights at 8 PM, with each session lasting 30 minutes. In the program, Li Xiang will address three focal questions from netizens: whether car manufacturers can engage in AI, Ideal Automotive's understanding of intelligent driving, and his personal reasons for buying a Ferrari.

In recent years, large language models represented by ChatGPT have sparked a new wave in the AI field. These powerful language models can not only understand human instructions but also formulate plans, explore environments, and utilize tools to solve complex tasks, demonstrating immense potential in areas such as robotics, personal assistants, and process automation. However, most existing AI entity systems are passive, requiring explicit human instructions to perform tasks. To schedule a meeting, one still has to manually input the time, location, and even list out all the participants, which is quite cumbersome.

The Massachusetts Institute of Technology (MIT) showcased a brand new robot training model this week that abandons the previous focus on training methods centered around specific datasets, opting instead for an approach that utilizes vast amounts of information similar to that used in training large language models (LLMs). Researchers noted that imitation learning—where agents learn by mimicking individuals performing tasks—can fail when facing minor challenges. These challenges may include varying lighting conditions, different environmental setups, or new obstacles. In such cases, the robots lack sufficient data to adapt to these changes.

This week, MIT showcased a new model for training robots, aimed at addressing the challenges faced by imitation learning when introduced to small obstacles. Researchers pointed out that imitation learning can fail in conditions such as varying lighting, different environments, or new obstacles because robots simply do not have enough data to adapt. The team sought powerful data methods, like those used in models such as GPT-4, to tackle the problem. They introduced a new architecture called the Heterogeneous Pretrained Transformer (HPT), which aggregates information from different sensors and environments.