Amazon acquires founders of AI startup Covariant for warehouse robot software. Deal includes licensing Covariant's AI models. Covariant Brain platform improves robot efficiency. Amazon plans to enhance automation with Covariant's technology. Other tech giants also recruiting AI startup talent and licensing technology.

The AI industry faces a potential fourth winter due to high project failure rates. "Intelligent engineering" offers a solution by combining scientific breakthroughs with engineering practices. Implementing five steps can introduce intelligent engineering and open a new era of AI value creation.

Artificial Intelligence (AI) revolutionizes database interactions with natural language queries. TAG technology bridges the gap between language models and database systems, enhancing query accuracy and speed. It offers a promising solution for complex data analysis without extensive coding.

In the fast-growing field of AI, domestic unicorn MiniMax has entered the video generation model competition. At "MiniMax Link Partner Day," founder Yan Junjie announced the latest models, including "video-1," with high compression and diverse styles. The company aims to enhance user creative experience and plans iterative upgrades. Despite the model's initial strength, there is room for improvement in facial detail handling. MiniMax's move into video generation aligns with the increasing relian

OpenAI enhances file search control in Assistants API, improving responses. Developers can now adjust how agents select information and configure search results ranking for relevance. This update addresses the need for fine-tuning assistants in building autonomous AI applications.

Multi-modal models face challenges in combining text and images. Transfusion, a new technique by Meta and USC, trains a single model with language and diffusion objectives to handle both modalities seamlessly. It outperforms quantization-based methods in various benchmarks.

Amazon's Alexa to run on Anthropic's chatbot Claude due to internal AI struggles. Amazon invested $4 billion in Anthropic, seeking enhanced AI for Alexa to handle complex requests. New generative AI named "Remarkable" to be charged monthly fee of $5 to $10. Apple also considering investment in Anthropic.

Cohere upgrades Command R and Command R+ AI models with improved RAG, multilingual capabilities, and enhanced tool usage. These models offer increased efficiency and cost-effectiveness in AI operations for enterprises. The new features include better multilingual support, optimized programming capabilities, and Safety Modes for flexible governance. Available through Cohere's API and Amazon SageMaker. Command-r-08-2024: $0.15 per million input tokens, $0.60 per million output tokens; Command-r-pl

Kunlun Wanyi launched AI music platform "Yin Feng" for easy music creation. Features include AI song creation, melody motifs, and diverse music styles. Users can upload and sell creations, choose membership packages. Suitable for beginners, musicians, content creators, social media users, educators. Be cautious with personal information security.

Nvidia released "Eagle," a new AI model advancing multimodal large language models with enhanced visual processing capabilities. It excels in OCR, visual question answering, and document understanding tasks, showing promise in various industries like law, finance, and e-commerce. Nvidia's open-sourcing of Eagle aims to drive collaboration and transparency in AI research while emphasizing ethical AI development. Eagle's performance may lead to new applications in assistive technologies, content m

Norwegian startup 1X launches humanoid robot NEO Beta for home use. Lightweight & quiet, NEO mimics human gestures & has advanced AI "Embodied Intelligence." NEO's nearly silent operation & fine motor skills make it a strong competitor in the humanoid robot market.

OpenAI's ChatGPT now boasts 200 million weekly active users, doubling since last year. Fortune 500 companies are actively adopting OpenAI products. ChatGPT's integration with Siri in iOS 18 promises to enhance user experience without direct financial transactions.

Google researchers developed AI "GameNGen" that mimics Doom gameplay using neural networks. GameNGen generates new game frames based on previous ones, creating coherent content. The model, though not flawless, impressively constructs levels dynamically around players. The AI relies on vast amounts of game data and trained agents playing Doom to simulate human gameplay. Scientists use Gaussian noise and image feedback loops to improve image generation accuracy. GameNGen is a significant step in A

IBM will introduce Intel's Gaudi 3 AI processor on its cloud platform, offering improved energy efficiency and training time compared to Nvidia's H100 GPU. Gaudi 3 features enhanced processing modules and a high-speed memory pool, with support for interconnectivity via a built-in Ethernet module.

Alibaba Cloud unveiled Qwen2-VL, a visual language model enhancing visual understanding, video analysis, and multilingual text-image processing. The model rivals top models like Meta's Llama 3.1 and OpenAI's GPT-4, offering powerful visual analysis, video content understanding, and multiple version options. Its architecture optimizations enable integration with third-party software for real-time operational scenarios. Explore its potential through the Hugging Face platform.

Meta's Llama models gain popularity, reaching new heights with over 350 million downloads on Hugging Face. Open-source strategy enables Meta to compete with leaders like OpenAI, attracting big enterprises. Closed-source AI providers face challenges as open-source AI performance advantages become evident. OpenAI cuts prices in response, but innovation pace lags behind.

NVIDIA surpasses revenue expectations, reaching $30.04 billion with a 122% YoY increase, driven by AI investment boom. CEO Huang anticipates record chip production in 2022. Stock prices decline due to growth concerns and production issues, amid warnings of AI bubble burst.

OpenAI is seeking massive investment, with potential involvement from Apple, Nvidia, and Microsoft. Apple plans to integrate ChatGPT into iOS, while Nvidia supplies chips for OpenAI's AI services. Microsoft has previously invested in OpenAI.

Google has launched a Gmail Q&A feature on mobile devices, enabling users to ask questions about their inbox. The feature, powered by Gemini AI, offers efficient email filtering and searching, but requires subscription to Google One AI Premium or certain Google Workspace plans.

Google's research team created GameNGen, a neural network system for real-time gameplay in "Doom" without traditional game engines. GameNGen revolutionizes AI technology in game development, predicting high-quality graphics at 20 frames per second. GameNGen, using Google's TPU AI chip, enhances game simulation and may transform the $200 billion gaming industry, reducing costs and enabling new game genres. It also offers potential applications in virtual reality, autonomous driving, and smart cit