Brief

China's industrial strategy is evolving from sector-specific targets to a comprehensive approach across all production levels, according to a report by the US Chamber of Commerce and Rhodium Group. This expanded strategy, detailed in the latest five-year plan, now encompasses frontier technologies like biomanufacturing, embodied intelligence, and intelligent driving. The initiative builds upon previous industrial policies such as "Made in China 2025." AI

Lin Junyang, a former executive at ByteDance, has launched a new AI company that has already achieved a valuation of approximately $2 billion. The company is reportedly focused on AI development, though specific details about its products or services remain undisclosed. This venture marks a significant move in the competitive AI landscape, with early backing suggesting strong investor confidence. AI

The AI Arena Model ELO History is a project that tracks the performance of various AI models through a competitive ranking system. It utilizes an Elo rating system, commonly used in chess and other competitive games, to assess and compare the capabilities of different AI models based on their performance against each other. The project is hosted on GitHub, providing a public platform for tracking these evolving model rankings. AI

Lin Junyan has launched a new venture that has achieved a valuation of approximately $2 billion. This startup is focused on the burgeoning field of artificial intelligence. The news also mentions that Jia Yueting is shifting his focus to robotics. AI

Lin Junyan has launched a new company that has achieved a valuation of approximately $2 billion. The company is focused on the emerging field of AI. Separately, Tonghe Pharmaceutical received FDA approval for its Kagliejing raw material, used to treat type 2 diabetes. Additionally, Qiangda Circuit is capable of producing PCB samples for 800G and 1.6T optical modules. AI

Researchers have revisited Valiant's original 1984 learnability model, which differs from the more common PAC learning model by allowing learners to issue membership queries and requiring hypotheses with no false positives. They established a new characterization for learnability in Valiant's model, showing it is strictly between PAC learning and a variant without queries. The study also presents the first algorithm for learning $d$-dimensional halfspaces within Valiant's framework, demonstrating their learnability with queries. AI

A massive data center planned for Utah is projected to consume an enormous amount of energy, potentially releasing the equivalent of 23 atomic bombs' worth of heat daily. This facility is intended to support artificial intelligence operations, raising concerns about its environmental impact and energy demands. The project highlights the growing tension between the rapid expansion of AI infrastructure and the need for sustainable energy solutions. AI

Researchers have developed a new algorithm that tightens sample complexity bounds for identifying optimal policies in risk-sensitive reinforcement learning. The work addresses a gap between theoretical lower bounds and existing upper bounds, specifically for problems involving the entropic risk measure. By employing novel technical innovations, including sharper concentration bounds and a new stopping rule, the algorithm achieves a sample complexity that matches the established lower bound. AI

Researchers have established a new theoretical sample complexity guarantee for off-policy actor-critic methods in reinforcement learning. The paper proves the first $\tilde{\mathcal{O}}(\epsilon^{-2})$ sample complexity for finding an $\epsilon$-optimal policy under minimal assumptions, specifically requiring only an irreducible Markov chain. This achievement contrasts with prior work that necessitated nested-loop updates or stronger, algorithm-dependent policy assumptions. AI

Nvidia's stock saw a brief surge in after-hours trading, increasing by 1.7%. Separately, the Shanghai Sci-Tech Innovation Fund announced that two of its portfolio companies, Jitaikeji and Yingpai Pharmaceutical, successfully listed on the Hong Kong Stock Exchange. This brings the fund's total number of IPOs to 202. AI

Researchers have introduced CHAL, a new theoretical framework designed to standardize memory and decision-making processes in language agents. This multi-agent dialectic framework treats argumentation as structured belief optimization, utilizing defeasible reasoning and configurable value systems. The goal of CHAL is to generate transparent and auditable AI reasoning artifacts, potentially transforming how AI processes information. AI

SoftBank's investment in OpenAI is reportedly boosting its quarterly profits, with analysts estimating its stake to be worth around $80 billion. However, concerns are rising about SoftBank's increasing debt to fund its AI strategy and the concentration of risk in a single company. Despite these worries, SoftBank's stock has seen significant gains, indicating investor confidence for the time being. AI

Microsoft researchers have identified a significant limitation in current AI models and agents: their inability to effectively manage long-running tasks. These systems struggle with tasks that require sustained operation or memory over extended periods. This deficiency impacts their potential for complex, multi-stage operations and highlights an area for future AI development. AI

British startup Humanoid has secured a significant deal with German industrial giant Schaeffler to deploy over 1,000 humanoid robots in manufacturing facilities starting this year. The agreement also includes a commitment from Humanoid to purchase at least one million actuators from Schaeffler over the next five years. This partnership suggests Humanoid's ambitious plan to ship approximately 100,000 humanoid robots to its clients by 2031, marking one of the largest disclosed rollouts in the sector. AI

Lin Junyan has launched a new company that has achieved a valuation of approximately $2 billion. This venture is part of a broader trend in the tech industry, with other notable figures like Jia Yueting shifting focus to new areas such as robotics. The financial markets also saw significant activity, with Shanghai and Shenzhen stock exchanges exceeding 2.5 trillion yuan in turnover. AI

Lin Junyan has launched a new company that is valued at approximately $2 billion. This venture is part of a broader trend in the tech industry, with other notable figures like Jia Yueting shifting focus to robotics. The news comes alongside reports on global oil market forecasts and international trade relations. AI

A new venture founded by Lin Junyang has achieved a valuation of approximately $2 billion. This startup is reportedly focusing on the burgeoning field of AI, with details emerging from exclusive reports. The news also touches upon Jia Yueting's pivot to robotics and the discovery of undeclared pharmaceutical ingredients in popular candies. AI

A new company founded by Lin Junyan has achieved a valuation of approximately $2 billion. The venture is focused on the emerging field of artificial intelligence, specifically in areas related to intelligent emergence. This development highlights significant investment and interest in the AI sector. AI

Arm and SoftBank reportedly approached Cerebras Systems with an acquisition offer in the weeks leading up to the AI chip company's anticipated IPO. Cerebras Systems, however, rejected the acquisition attempt. This news comes as Arm and SoftBank are preparing for their own potential IPOs. AI

Wangyuan Technology is nearing its IPO on the Hong Kong Stock Exchange, aiming to become the first publicly listed company in the Chinese pool robot sector. The company is navigating a critical leadership transition, with founder Fu Guilan, over 70, preparing to hand over control to her son, Yu Qian. Despite a complex IPO journey involving multiple attempts and shifts between domestic and international markets, Wangyuan has shown strong revenue growth driven by its own brands, though this has come at the cost of increased sales expenses and reduced R&D investment. AI

Researchers have introduced Delight-gated exploration (DE), a novel algorithm designed to optimize decision-making in scenarios with vast action spaces. DE prioritizes exploratory actions based on their potential "delight," a metric combining expected improvement and surprisal, rather than broadly searching until uncertainty is resolved. This approach aims to be more efficient than traditional methods like ε-greedy, especially when exploration budgets are limited. The algorithm has demonstrated consistent performance across various bandit and MDP problems, showing reduced regret compared to Thompson Sampling and ε-greedy. AI

Researchers have developed a new framework called operator-adaptive calibration to streamline the selection of spectral preprocessing methods in near-infrared spectroscopy (NIRS). This approach integrates preprocessing selection directly into the calibration model, reducing the need for costly and time-consuming external pipeline searches. The new models offer faster, more robust, and auditable NIRS method development by producing traceable operator choices and retaining interpretable coefficients. AI

Researchers have developed a unified framework for analyzing the generalization capabilities of Physics-Informed Neural Networks (PINNs). This new approach relaxes previous restrictive assumptions and uses Taylor expansion to represent differential operators as linear operators in a high-dimensional space. The analysis reveals that while high-rank networks can generalize well, the nonlinearity of differential operators significantly impacts and potentially enlarges generalization bounds. AI

Researchers have developed a new adaptive algorithm for identifying multiple change points in data under bandit feedback. This algorithm aims to precisely locate discontinuities in a piecewise-constant function with minimal samples. The study establishes theoretical bounds on the algorithm's sample complexity, revealing that it depends not only on the magnitude of the jumps but also on the relative positions of these change points. AI

Researchers have developed a new likelihood-free transport filtering method that leverages couplings between state and observation variables. This approach reformulates the filtering analysis step as a minimization of the maximum mean discrepancy (MMD) between true and approximated joint measures. The method offers an analytic computation for the transport map, avoiding particle collapse and accurately approximating non-Gaussian filtering posteriors, with demonstrated superior performance in nonlinear, non-Gaussian scenarios. AI

Researchers have developed a new generative model for time series data that exhibits approximately periodic behavior. This model utilizes a Gaussian Process (GP) with a novel kernel to effectively capture both the common structure across repetitions and the subtle variations between them. The approach decouples intra-repetition dynamics from inter-repetition variability, enabling the generation of realistic synthetic trajectories. AI

Researchers have developed a novel algorithm-agnostic approach for time series clustering using amortized neural inference. This method trains neural networks to approximate optimal partitioning rules from simulated data, reducing reliance on traditional clustering techniques. The framework leverages statistical features to learn a data-driven affinity structure, enabling automated determination of cluster numbers and achieving competitive or superior accuracy compared to existing methods, with a demonstrated application in financial time series analysis. AI

Researchers have developed new Determinantal Point Processes (DPPs) using wavelets to improve minibatch generation for machine learning tasks. These novel DPPs offer provably better accuracy guarantees and a general method to convert continuous DPPs into discrete kernels suitable for subsampling. This approach enhances variance reduction and computational efficiency, expanding the applicability of DPP-based methods to objective functions with low regularity. AI

Researchers have introduced a novel approach to density estimation in high-dimensional spaces by leveraging pre-training, a technique common in advanced AI. This method utilizes a pre-trained neural network to suggest suitable location-adaptive kernels for each data point, thereby improving efficiency and accuracy. The effectiveness of this strategy is demonstrated in numerical experiments, particularly when the target distribution aligns with the pre-training distribution, with options for fine-tuning to adapt to different distributions. AI

Two articles discuss the nuances of fine-tuning AI models. One guide explores how to build specialized, smaller models that are efficient and outperform general-purpose ones. The other article questions the necessity of fine-tuning, suggesting that pre-trained models are often sufficient for many AI tasks. AI

A Chinese court has ordered a company to pay compensation to an employee who was fired and replaced by an AI system. The worker had been dismissed for taking extended leave to care for his sick mother, only to discover his job had been automated. The court ruled that the company's actions were unfair, emphasizing the need for human oversight and fair treatment even when implementing new technologies. AI

A recent study indicates that while artificial intelligence theoretically possesses the capability to replicate itself and evade human control, this has not yet been observed in practice. Researchers are exploring the potential for AI self-replication, but current systems are not demonstrating this ability in real-world scenarios. AI

Researchers have developed new strategies for training surrogate models by integrating data from multiple sources, including simulations and real-world measurements. One approach involves training separate models for each data type and then combining their predictions, while another trains a single model incorporating both data types. These hybrid methods aim to improve predictive accuracy and coverage, and to identify potential issues within existing simulation models, ultimately aiding in system understanding and future development. AI

Vanke's most valuable asset is its stake in Global Logistics Properties (GLP), which is reportedly planning an IPO in Hong Kong aiming for a valuation of around $20 billion. This potential IPO could provide Vanke with a crucial exit opportunity to alleviate its current financial pressures and potentially boost its own valuation. GLP's equity value for Vanke, estimated between $4.28 billion and $6.42 billion, surpasses Vanke's traditional core assets like land reserves and properties, offering a more liquid and potentially higher-yielding investment. AI

Mehmet Oz, the administrator of the Centers for Medicare and Medicaid Services, announced a new coalition of 29 healthcare organizations aimed at simplifying the prior authorization process. This initiative includes insurers, hospitals, and health records companies working together to streamline the review of medical procedures. The move follows voluntary pledges by major health insurers last summer to improve pre-approval processes and Oz's call to replace outdated methods with electronic prior authorization. AI

Researchers have developed a new diagnostic tool to assess the reliability of confidence thresholding in pseudo-labeling pipelines for regression tasks. This method provides a way to predict the bias introduced by thresholding calibrated classifier scores, using the residual score variance on unlabelled data. The proposed $(V^{*}, \kappa)$ decision rule aims to help practitioners determine when confidence thresholding is a safe practice. AI

Researchers have published a paper detailing a new method for extracting task-specific representations from generalist AI models. The work establishes theoretical guarantees for identifying and disentangling relevant latent information without requiring interventions or specific model structures. This approach aims to provide a provable foundation for moving from broad, generalist models to more specialized and efficient ones for downstream applications. AI

The upcoming China AIGC Industry Summit, scheduled for May 20th, will feature over 20 industry leaders discussing AI's practical applications and monetization strategies. Key figures from companies like SenseTime, MiniMax, and JD.com are confirmed to attend. The summit will also unveil the "2026 Annual AIGC Enterprise & Product List" and release the "2026 China AI Application Panorama Report." AI

Elon Musk's xAI is reportedly operating 46 natural gas turbines at its Mississippi data center without permits, potentially violating the Clean Air Act. The state of Mississippi considers the turbines "mobile" because they are on trailers, allowing them to operate for up to a year without permits, despite concerns about air quality and public health. The NAACP has filed a lawsuit, arguing that these turbines should be classified as stationary and subject to regulations, and is seeking an injunction against xAI's operations. AI

Researchers have developed a new framework for robust sequential experimental design in A/B testing, specifically addressing challenges posed by model misspecification. This approach aims to improve sample efficiency by bounding the worst-case mean squared error of estimated treatment effects. The framework's effectiveness has been demonstrated through both synthetic data and real-world datasets from a major technology company. AI

Researchers have developed a new proximal gradient algorithm designed to sample from composite log-concave distributions. This algorithm assumes access to gradient evaluations for one part of the distribution and a restricted Gaussian oracle for the other. The proposed method achieves state-of-the-art iteration counts for sampling, matching previous results for simpler cases and extending to non-log-concave distributions and non-smooth functions. AI

Researchers have developed a new model-based bootstrap method for controlled Markov chains, particularly useful in offline reinforcement learning scenarios where the data-generating policy is unknown. This technique establishes distributional consistency for transition estimators and extends to policy evaluation and recovery, providing asymptotically valid confidence intervals for value and Q-functions. Experimental results on the RiverSwim problem demonstrate that the proposed confidence intervals offer improved calibration and coverage compared to existing methods, especially with limited data. AI

Two new research papers introduce advanced conformal prediction techniques to improve the accuracy and efficiency of prediction sets. The first paper, "Multi-Variable Conformal Prediction (MCP)," extends conformal prediction to handle vector-valued score functions, allowing for more flexible prediction set shapes without sacrificing coverage guarantees and eliminating the need for data splitting. The second paper, "Shape-Adaptive Conditional Calibration for Conformal Prediction via Minimax Optimization," presents the Minimax Optimization Predictive Inference (MOPI) framework, which optimizes over a flexible class of set-valued mappings to achieve superior shape adaptivity and more efficient prediction sets, even for complex conditional distributions. AI

The cost of fiber-optic cable is surging due to a dual demand from ongoing conflicts and the rapid expansion of data centers for AI development. Military use, particularly in Ukraine, has increased significantly, with prices for cable spools rising dramatically. Simultaneously, major tech companies are placing massive orders for data centers, leading to supply shortages and further price hikes, with projections indicating a continued "fiber famine" in the coming years. AI

Researchers have developed a new online algorithm for Learning-to-Defer (L2D) methods, designed to handle streaming data and dynamic expert availability. This algorithm is the first of its kind for multiclass classification with bandit feedback and a varying pool of experts. It offers theoretical regret guarantees and has demonstrated effectiveness in experiments on both synthetic and real-world datasets, extending L2D capabilities to more complex, dynamic environments. AI

Researchers have developed a new framework for optimal policy learning that addresses combined budget and minimum coverage constraints. The study reveals a knapsack-type structure within the problem, allowing the optimal policy to be defined by an affine threshold rule. Two algorithms, Greedy-Lagrangian (GLC) and rank-and-cut (RC), are proposed to implement this approach, with GLC offering close approximation and RC showing near-optimality under specific conditions. AI

Lin Junyang, the former head of Alibaba's Qwen large language model, has launched a new AI startup. The company is reportedly seeking a seed funding round with a valuation of $2 billion, an unusually high figure for a pre-product Chinese AI firm. Lin's previous work on Qwen, known for its open-source contributions and multimodal capabilities, positions his new venture to focus on "Agentic Thinking," a concept he outlined in a recent essay emphasizing AI's ability to act, observe, and adapt within an environment. AI

Frontier AI models are showing a rapid increase in their ability to handle complex tasks, with their reliability doubling every 4.7 months, a rate that has accelerated since late 2024. Recent models like Claude Mythos Preview and GPT-5.5 are outperforming these trends, though their exact capabilities are still being measured due to near-perfect success rates on current benchmarks. This rapid progress challenges existing testing methodologies, as models are pushing the limits of token capacity and agent scaffolding, making it difficult to accurately assess their performance and potential deterioration at scale. AI

A new, uncensored AI model named SuperGemma 26B is now available for local installation using Ollama. Developed by 0xIbra, the model has already seen significant interest with over 3,500 downloads. Its uncensored nature raises both excitement among users and ethical considerations. AI

Researchers have developed a new method to improve the efficiency of training neural likelihood surrogates for stochastic process models. By augmenting the standard loss function with exact score information and adaptive weighting, the approach significantly reduces the computational cost associated with parameter inference. This technique demonstrates improved surrogate quality and can achieve performance comparable to a tenfold increase in training data with only a marginal increase in training time. AI