LingBot-VLA 2.0, the ant-like Lingbo embodiment base model, is open source.
Related
Zhang Yuhang, Beijing Municipal Bureau of Economy and Information Technology: Developing the word-based economy and exploring business models such as quantity-based pricing, performance-based payment, and subscription-based hosting.
According to Mars Finance, Zhang Yuhang, a second-level inspector of the Beijing Municipal Bureau of Economy and Information Technology, stated on July 8th that the next step will be to create a new form of intelligent economy. This includes developing a word-based economy, building word-based factories, and exploring business models such as quantity-based pricing, performance-based payment, and subscription-based hosting; accelerating research and development of key technologies for embodied intelligence and robot products, and improving the systematic evaluation capabilities of robots; supporting the innovative development of intelligent terminals such as smartphones, smart computers, smart glasses, and smart homes, and improving the pilot-scale testing service system. (Beike Finance)
AI giants like OpenAI and Anthropic are offering startups large amounts of free computing power in an effort to seize market share.
According to a report by the Wall Street Journal on July 7th, as reported by Mars Finance, AI companies such as OpenAI and Anthropic are offering substantial free computing power and discounts to startups to compete for enterprise customers. The report states that Silicon Valley startup founders are receiving computing credits, token usage limits, and auction-style discounts from AI model companies. Some early-stage companies have received over $3 million in cloud computing and token credits, approaching the median of seed funding in the US according to PitchBook. AI companies hope to acquire customers in the early stages of startups, making their tools an integral part of their business as these companies grow. Cursor offered a 75% discount until July 5th; Google Cloud offered up to $500,000 in cloud computing credits to some startups, along with early access to the Gemini model and, in some cases, support from DeepMind engineers. Microsoft and Amazon Web Services also offer special privileges to startups. OpenAI and Anthropic have recently been particularly focused on Y Combinator startups. In May, Sam Altman announced that OpenAI would offer $2 million in tokens to each startup participating in its accelerator program in exchange for equity. Around the same time, Anthropic increased its free token allocation to Y Combinator startups from $30,000 to $500,000, without requiring equity. OpenAI subsequently adjusted its offer, providing startups with $500,000 in free tokens, without requiring equity, and offering the option to exchange equity for an additional $1.5 million in tokens. These offers reflect the fierce competition among model providers for future large clients. Y Combinator runs four cohorts annually, with recent cohorts featuring approximately 200 companies each, meaning that OpenAI and Anthropic could potentially offer a combined $800 million in AI tokens over the next year. Christopher Acker, co-founder of SuperPenguin, stated, "The AI world is being driven by OpenAI and Anthropic because they are giving startups money to pay for usage costs."
MIRA, an open-source 5B multiplayer world model, uses DINOv3 representation to mitigate long-term drift and can simulate 2v2 battles in Rocket League in real time.
According to Beating, AI research firm General Intuition, in collaboration with French AI lab Kyutai and Epic Games, has launched MIRA, a multiplayer interactive world model. As a generative game simulator supporting real-time multiplayer interaction, MIRA can simulate 2v2 battles in Rocket League in real time, based solely on historical footage and player button presses, without requiring a physics engine, rendering engine, or explicit 3D representation for inference. Unlike the "decoupling of logic computation and image rendering" approach adopted by companies like Odyssey, MIRA takes a generative simulation approach based on video latent space. MIRA boasts 5 billion parameters, and its core design builds the latent prediction space on a frozen general-purpose visual encoder, DINOv3-L. Leveraging pre-trained visual features, the generated latent states can more stably fall within the effective representation space, significantly mitigating image drift and divergence during long-term prediction. For multi-screen alignment, MIRA stitches the latent images from four player perspectives into a unified grid, enabling spatial attention mechanisms to operate naturally across viewpoints and improving the spatial consistency of vehicles, the soccer ball, and key events across multiple perspectives. The Action Dropout introduced during training also helps the system complete the game behavior of vehicles not controlled by commands when parts of the motion flow are missing. Currently, MIRA can run in real time at 20 frames per second on a single NVIDIA B200 graphics card. The team has open-sourced the training and inference code and released the Rocket Science dataset, which contains 1,000 hours of matches, approximately 4,000 hours of video, motion flow, and physics data from four perspectives; the complete training of the model used approximately 10,000 hours of clean match data.
Inside Coinbase’s $250 Billion Playbook for Post-Quantum Bitcoin Custody
Head of cryptography Yehuda Lindell says the exchange is designing custody that can adapt to whatever post-quantum signing scheme Bitcoin adopts.
Ministry of Human Resources and Social Security: By 2030, the large-scale model of the human resources and social security industry will be mature and complete, and the artificial intelligence application system will be basically sound.
According to Mars Finance, the Ministry of Human Resources and Social Security, the National Development and Reform Commission, the Ministry of Industry and Information Technology, and the National Data Administration recently jointly issued the "Implementation Opinions on Accelerating the Application and Development of 'Artificial Intelligence + Human Resources and Social Security'". A relevant official from the Ministry of Human Resources and Social Security answered reporters' questions regarding the "Implementation Opinions". In terms of promoting the work, a three-step work goal will be achieved within five years. First, building the foundation. Taking this year (2026) as a benchmark, the initial formation of the artificial intelligence application system, standard system, and guarantee system in the human resources and social security sector will be promoted. The infrastructure for "Artificial Intelligence + Human Resources and Social Security" applications will be deployed, a number of high-performance human resources and social security industry large-scale models and intelligent agent applications will be cultivated, and about 20 application scenarios based on human resources and social security industry large-scale models and corresponding high-quality datasets will be created, forming a collaborative development ecosystem of computing power, models, data, and scenario applications. Second, popularization and promotion. By 2027, a number of human resources and social security industry large-scale models and intelligent agents will be widely applied, and about 50 high-value application scenario empowerment paths will be explored, achieving significant results in intelligent development. Third, widespread application. By 2030, high-quality datasets will be effectively supplied, large-scale models for the human resources and social security industry will be mature and complete, and the artificial intelligence application system will be basically sound, forming an innovative landscape where artificial intelligence is widely applied in the human resources and social security sector. (Cailian Press)
BNB Agent Studio now supports the B402 merchant pool via Binance Pay.
According to official sources, Odaily Agent Studio now supports AI Agents to access CoinMarketCap (CMC) market data with one click through Binance Pay's B402 merchant pool, and the Agent's own wallet can automatically complete the data retrieval and payment. Previously, building an AI Agent that relied on real-time market data typically required configuring API keys, subscription plans, and separate payment processes. With this integration, developers no longer need to create CMC accounts, manage API keys, or write additional payment code; the Agent can complete the entire process of "requesting data—automatic payment—processing output." Currently, four types of CMC data interfaces are initially open, including: DEX Search: Search for tokens and trading pairs on decentralized exchanges; Quotes Latest: Get the real-time price and market data of a specified token; Listings Latest: Gets ranking data for tokens across the entire market; DEX Pairs Quotes: Get real-time quotes for a specified DEX trading pair. BNB Chain states that all data requests are processed via B402 based on the x402 protocol and settled instantly on the BNB Smart Chain (BSC). This feature will help developers quickly build AI Agents with market awareness capabilities, such as daily market analysis agents, token ranking change alert agents, investment watchlist agents, and narrative analysis agents that combine large language models for market interpretation.