Recent research from Alibaba's Tongyi Laboratory indicates that existing text-to-image Diffusion Transformer models already possess the ability to generate multiple images with specific relational contexts. With just a little 'guidance', they can achieve a synthesis of understanding and produce high-quality multi-image collections. Traditional Diffusion models are more like 'rote learning' students, requiring massive datasets to generate high-quality images. However, with the support of IC-LoRA, they become 'high-achieving students', capable of learning through analogy.

Xiaomi's upcoming Xiaomi 15 series smartphones will no longer offer an 8GB RAM version, instead, the base model's RAM configuration will start at 12GB, with a projected price increase. Xiaomi representative Wang Teng noted that future flagship models are likely to eliminate the 8GB RAM option due to the higher memory demands of on-device AI models. He suggested that as AI technology becomes more widespread in smartphones, users should opt for higher RAM versions whenever financial conditions permit. He also advised consumers to be aware of the pricing differences of RAM across different brands.

Large Language Models (LLMs) like the GPT series exhibit astonishing abilities in language understanding, reasoning, and planning, achieving human-level performance in various challenging tasks due to their vast datasets. Most research focuses on enhancing these models further by training them on larger datasets, aiming to develop more powerful foundational models. However, while training stronger foundational models is crucial, researchers believe that empowering models with the capability to continue evolving during the reasoning phase, known as AI self-evolution, is essential for the development of AI.

White hat hacker Johann Rehberger recently revealed a shocking security vulnerability, finding that ChatGPT can be exploited by hackers as spyware, continuously stealing user data. Rehberger named this attack method "SpAIware," primarily targeting the new persistent memory feature of the macOS version of the ChatGPT application. This vulnerability exploits the recently launched persistent memory feature of ChatGPT. Through carefully crafted prompt injections, hackers can manipulate ChatGPT...

Recently, security researcher Johann Rehberger discovered a vulnerability in ChatGPT that could allow hackers to implant false information and malicious instructions in users' long-term memory. Although he reported the issue to OpenAI, unfortunately, the company did not take it seriously and quickly closed the related investigation, claiming it was not a security concern. Faced with this situation, Rehberger decided to persevere and developed a proof-of-concept attack example exploiting this vulnerability.

MIT research reveals that AI, particularly chatbots, can implant "false memories" in the human brain. Experiments show that AI can lead volunteers to develop persistent incorrect memories through questioning. Individuals interested in but unfamiliar with AI are more susceptible to this influence. This finding poses challenges for AI applications in law, healthcare, and education, emphasizing the need to ensure AI acts as a tool for protecting and....

MianBi Intelligence has recently launched MiniCPM 3.0, a high-performance AI model that operates on devices with just 2GB of memory, heralding a new era of edge AI experience. The MiniCPM 3.0 model features 4 billion parameters and outperforms GPT-3.5, providing AI services on mobile devices comparable to those of GPT-3.5. This allows users to enjoy fast, secure, and feature-rich local AI services without relying on cloud processing, ensuring a smoother and more private intelligent interaction experience.

Magic, a startup focused on generative AI programming, has announced the completion of $320 million in funding, bringing its total financing to nearly $465 million. This round of funding was led by former Google CEO Eric Schmidt, with participation from several well-known investors. Magic's core technology, the Long-Term Memory Network (LTM), can handle vast amounts of data and avoid bias when generating code. Its latest model, LTM-2-mini, can process 100 million tokens, equivalent to 10 million lines of code or 750 novels. In partnership with Google Cloud, it plans to establish two '

The upcoming iPhone 17 series from Apple is expected to feature 12GB RAM to meet the memory demands of future AI large models. The current iPhone 15 Pro and iPhone 16 series with 8GB RAM are insufficient for supporting large models with around 7 billion parameters. A model with 13 billion parameters requires at least 5GB of memory, and after the operating system and applications take up space, 8GB RAM fails to provide enough room. The 12GB RAM configuration could offer ample space for larger AI models.

A research team from Fudan University has developed the 'Concept Erasure Technique' (RECE), which can efficiently and accurately modify inappropriate ideas in AI models, significantly reducing the likelihood of generating inappropriate images while maintaining the AI's other creative capabilities. This technology uses a 'closed solution' to locate and modify specific parts, ensuring the AI behaves more appropriately during creation and stays on track. Experiments show that the RECE technology can effectively enhance the morality and quality of AI creations, providing new directions for the future development of AI, making future AI assistants smarter and more perceptive.

Samsung Electronics has started mass production of the industry's thinnest low-power Dynamic Random Access Memory (DRAM) chips to meet the growing demand for artificial intelligence (AI) in mobile devices. The newly launched 12-nanometer, 12GB, and 16GB LPDDR5X DRAM chips are only as thick as a fingernail, achieved through Samsung's expertise in chip packaging technology. This design not only provides more space for mobile devices but also improves airflow and optimizes thermal management. In terms of performance, Samsung's LPDDR5X DRAM chips deliver.

Ampere Inc. has announced a significant update to its product roadmap with the launch of its flagship processor, Aurora, designed for AI acceleration and high bandwidth memory integration, aiming to greatly enhance cloud-native AI computing performance. Aurora features 512 Ampere cores, offering more than three times the performance of existing processors, specifically engineered for AI training and inference workloads, demonstrating Ampere's ambitions in the AI computing space. Aurora utilizes scalable AmpereOn technology.

Zyphra has launched the Zamba2-2.7B language model, which is a milestone in the small language model domain. Its performance and efficiency have significantly improved, with a training dataset size of around 30 trillion tokens, reducing resource requirements during inference and making it an efficient solution for mobile applications. Key highlights include a twofold increase in response generation speed, a 27% reduction in memory usage, and a 1.29 times decrease in generation latency, particularly suited for applications requiring real-time interaction, such as virtual assistants and chatbots.

Meta has trained its latest AI model, Llama 3.1, using 16,384 GPUs, showcasing the astonishing speed of AI technological advancement. However, during this process, an average of one failure occurs every 3 hours, totaling 419 failures, with approximately half related to the H100 GPUs and their HBM3 memory. This data reveals the reliability challenges faced by supercomputing systems in their pursuit of performance breakthroughs. The complexity of the Llama 3.1 training cluster is comparable to a small city's neural network, with frequent failures. The Meta team has implemented strategies such as reducing...

Tencent's MetaGen Image Generation Large Model (Hunyuan DiT) has recently been upgraded to a 6GB VRAM version, making it easy for personal computer users to run. This version is compatible with plugins such as LoRA and ControlNet and has added support for the Kohya graphical user interface, lowering the threshold for developers to train personalized LoRA models. The Hunyuan DiT model has been upgraded to version 1.2, with improvements in image texture and composition.At the same time, Tencent ha

Microsoft is set to introduce AI Explorer as PC manufacturers begin launching AI PCs. Microsoft will release a significant Windows 11 update that includes the 'AI Explorer' feature. AI Explorer is the evolved version of Microsoft's AI assistant and will debut in the Windows 11 24H2, scheduled for release this fall.

The KOALA AI model has been released and can run on 8GB memory, generating high-quality images in just 2 seconds. Developed by a South Korean scientific team, this new artificial intelligence image generation model uses 'knowledge distillation' technology, reducing parameters to 700 million, enabling image generation with only 8GB of memory. The generation time has been shortened to 2 seconds, making the KOALA model faster than OpenAI's DALL-E model.

Micron Technology has achieved revenue growth for six consecutive quarters, primarily due to the surge in demand for AI chips. The company's newly launched HBM3E memory chip modules have been well received and are expected to generate hundreds of millions in revenue for the company in 2024. Micron is optimistic about the applications of AI in personal computers and mobile devices, planning to increase capital expenditures in fiscal year 2024 to meet production demands for the new memory modules. The CEO stated that the company’s HBM supply is sold out and predicts a significant increase in sales in the personal computer sector in 2024, driving demand for memory and storage.

["Samsung confirms that the codename for the fifth generation HBM3E memory is Shinebolt, with a 50% bandwidth increase to 1.228TB.","Shinebolt will stack 24Gb chips in 8 layers, and a 12-layer 36GB product is under development.","HBM3E will provide strong support for the next generation of AI GPUs.","Samsung accelerates HBM3E development, closely following industry leader SK Hynix.","Samsung announces plans to further expand HBM production capacity to offer customized services to customers."]