但论文作者表示，至少对华尔街而言，Kalshi相比传统预测方式可能具备诸多优势。该平台上的合约交 易活跃，能针对更广泛的变量产出实时更新结果，同时更能全面反映各类潜在结果的可能性。 西北大学凯洛格管理学院博士研究生、论文联合作者贾里德・迪恩・卡茨（Jared Dean Katz）表 示："其真正的优势在于，能给出结果的概率分布而非单一的点估计，还能观察到事件发生后市场的快 速反应。" 他补充道："更难得的是，这些预测的准确率相当高。" 与竞争对手一样，Kalshi推出的各类"是/否"合约均以每份1美元兑付。例如，若某份合约售价32美分， 即意味着市场认为该结果发生的概率为32%。针对经济数据的合约，大多与消费者价格指数（CPI）等 最终数据是否高于某一特定水平挂钩。 论文指出，从2022年至2026年6月的数据分析来看，Kalshi平台上的最可能预测结果（即交易者认为发 生概率最高的结果），在美联储决议公布前一晚的预判均完全准确。尽管美联储通常擅长引导市场预 期，但在2024年9月其推出超预期的50个基点降息时，该预测市场的表现依然亮眼。 一项新研究显示，炙手可热的预测平台Kalshi涉足了从超级碗赛事到美国 ...

但如果把问题继续往深处追问：推理能力的本质，真的只是多算几步吗？ 谷歌、芝加哥大学等机构的研究者最近发表的一篇论文给出了一个更具结构性的答案，推理能力的提升并非仅源于计算步数的增加，而是来自模型在推理 过程中隐式模拟了一种复杂的、类多智能体的交互结构，他们称之为「思维社会」（society of thought）。 过去两年，大模型的推理能力出现了一次明显的跃迁。在数学、逻辑、多步规划等复杂任务上，推理模型如 OpenAI 的 o 系列、DeepSeek-R1、QwQ- 32B，开始稳定拉开与传统指令微调模型的差距。直观来看，它们似乎只是思考得更久了：更长的 Chain-of-Thought、更高的 test-time compute，成为最常 被引用的解释。 简单理解就是，这项研究发现，为了解决难题，推理模型有时会模拟不同角色之间的内部对话，就像他们数字大脑中的辩论队一样。他们争论、纠正对 方、表达惊讶，并调和不同观点以达成正确答案。人类智能很可能是因为社交互动而进化的，而类似的直觉似乎也适用于人工智能！ 通过对推理输出进行分类，以及结合作用于推理轨迹的机制可解释性方法，研究发现，诸如 DeepSeek-R ...

Core Insights - The article discusses the significant leap in reasoning capabilities of large models over the past two years, highlighting the advancements made by models like OpenAI's o series, DeepSeek-R1, and QwQ-32B in complex tasks such as mathematics and logic [1][2] - It emphasizes that the improvement in reasoning ability is not merely due to increased computational steps but rather stems from a complex, multi-agent-like interaction structure termed "society of thought," where models simulate internal dialogues among different roles to arrive at correct answers [2][3] Group 1: Reasoning Mechanisms - The research indicates that reasoning models exhibit higher diversity of perspectives compared to baseline models, activating a broader range of features related to personality and expertise during reasoning tasks [2][3] - Controlled reinforcement learning experiments show that even with reasoning accuracy as the only reward signal, base models spontaneously increase dialogic behaviors, suggesting that socialized thinking structures enhance exploration of solution spaces [3][4] Group 2: Dialogic Behaviors - The study identifies four types of dialogic behaviors in reasoning trajectories: question-answer sequences, perspective shifts, viewpoint conflicts, and viewpoint harmonization, which collectively enhance cognitive strategies [7][8] - The Gemini-2.5-Pro model's evaluations show high consistency with human scoring, indicating reliable identification of these dialogic behaviors [9][13] Group 3: Social Emotional Roles - The analysis categorizes social emotional roles in reasoning trajectories into 12 types, which are further summarized into four high-level categories, demonstrating a balanced interaction among roles rather than isolated usage [10][22] - The Jaccard index is used to measure the co-occurrence of roles, revealing that models like DeepSeek-R1 organize different roles in a more coordinated manner during reasoning processes [10][22] Group 4: Cognitive Behaviors - The study identifies four cognitive behaviors that influence reasoning accuracy, including information provision, information inquiry, positive emotional roles, and negative emotional roles [11][12] - The consistency of the Gemini-2.5-Pro model's evaluations with human scoring reinforces the reliability of these cognitive behavior classifications [13] Group 5: Experimental Findings - The findings demonstrate that even with similar reasoning trajectory lengths, models exhibit a higher frequency of dialogic behaviors and social emotional roles, particularly in complex tasks [16][23] - Experiments show that guiding dialogic features positively impacts reasoning accuracy, with a notable increase from 27.1% to 54.8% in a specific task when dialogic surprise features are positively reinforced [24][29] Group 6: Reinforcement Learning Insights - A self-taught reinforcement learning experiment indicates that dialogic structures can spontaneously emerge and accelerate the formation of reasoning strategies when only correct answers are rewarded [30]

Core Viewpoint - The article discusses the rise of prediction markets, particularly Kalshi and Polymarket, as they gain mainstream attention during the 2024 U.S. presidential election, highlighting their potential as tools for forecasting events and the associated risks of being perceived as gambling platforms [4][6][20]. Group 1: Market Dynamics - Prediction markets have seen significant engagement, with over $3 billion wagered on election outcomes, proving to be more accurate than traditional polls [4][5]. - Kalshi and Polymarket have both surpassed $1 billion valuations, attracting substantial investments from venture capitalists [5][17]. - The platforms allow users to bet on a variety of events, from geopolitical issues to pop culture, indicating a broadening of market interests [4][5]. Group 2: Operational Mechanisms - Unlike traditional gambling, prediction markets operate on a peer-to-peer betting model, where participants bet against each other rather than against a house [8][9]. - The value of event contracts fluctuates based on public sentiment and market dynamics, providing real-time insights into the probability of various outcomes [9][10]. - Kalshi has established a more compliant operational framework compared to Polymarket, which has faced legal challenges [13][16]. Group 3: Challenges and Controversies - The industry faces scrutiny over its ethical implications, with concerns that it may be viewed as a form of gambling rather than a legitimate forecasting tool [6][20]. - Both platforms have encountered regulatory hurdles, with Polymarket previously banned from operating in the U.S. due to compliance issues [12][16]. - The perception of prediction markets as "casinos" could undermine their credibility and long-term viability [6][20]. Group 4: Future Prospects - The sustainability of interest in prediction markets beyond major events like presidential elections remains uncertain [7][18]. - New competitors are emerging in the prediction market space, indicating a potential shift in market dynamics [19]. - The platforms are exploring various revenue models, including transaction fees and partnerships with media and AI companies, to enhance profitability [19][20].

Core Viewpoint - The emergence of the "Ten Thousand Fund Era" in 2025 presents investors with unprecedented challenges in selecting from over 10,000 public funds, necessitating efficient decision-making tools and platforms [2][16]. Group 1: Selection Dilemma - The rapid increase in the number of public funds has led to a significant selection dilemma for investors, who must sift through vast amounts of information to identify potential investments [2][16]. - There are over 300 fund investment apps available, but very few effectively address the pain points faced by investors [2][16]. - The monthly active users of financial investment apps have surpassed 166 million, yet user retention rates exhibit a clear Matthew effect, indicating a disparity in information processing capabilities between professional and ordinary investors [2][16]. Group 2: Market Speed - The speed of market updates is crucial for investment success, with the Sina Finance app achieving a refresh rate of 0.03 seconds, significantly faster than the industry average [3][17]. - The app supports seamless connections to over 40 global markets, including A-shares, Hong Kong stocks, U.S. stocks, futures, foreign exchange, and precious metals [3][17]. - During critical market events, such as the May 2025 commodity futures night market crash, the app maintained millisecond-level updates, allowing users to capture cross-market arbitrage opportunities [3][17]. Group 3: AI Intelligence - The Sina Finance app addresses the challenge of data comprehension for ordinary investors through advanced data visualization and AI tools [4][18]. - The "Xina AI Assistant" can condense lengthy reports into concise summaries, highlighting risks and opportunities with color-coded indicators [4][18]. - The app's automatic strategy generation feature allows users to execute investment strategies based on real-time market conditions, enhancing decision-making efficiency [4][19]. Group 4: Trading Closure - The app integrates trading functionalities, providing a one-stop experience from information acquisition to transaction execution [5][20]. - Users can complete the entire process of account opening, fund transfer, and trade execution without switching apps, thanks to deep integration with over 40 major domestic brokers [5][21]. - The app's distributed trading gateway supports 120,000 concurrent transactions per second, maintaining zero latency during market fluctuations [5][21]. Group 5: Personalization - The Sina Finance app offers extensive personalization options, allowing users to customize their interface according to individual investment habits [6][23]. - The modular workspace supports over 200 functional components, enabling users to tailor their experience with various tools and features [6][23]. - The app's intelligent alert system monitors 12 types of conditions, achieving an alert accuracy rate of 98.2% [6][23]. Group 6: Aggregation and "Collective Wisdom" Effect - The app innovatively integrates insights from influential financial figures on Weibo, creating a dynamic loop of information, analysis, and trading [7][24]. - The community of certified analysts comprises 82% of participants, utilizing a keyword filtering system to eliminate noise and ensure high-quality discussions [7][25]. - The app's live broadcast sessions have reached up to 900,000 viewers, and research report downloads have exceeded 10 million, fostering a unique decision-making ecosystem [7][26].

Core Insights - The article discusses the relationship between intuition and collective intelligence, suggesting that intuition is not merely a personal gift but can be developed as a skill, especially in complex decision-making environments [1][2][3] Group 1: Intuition and Its Mechanisms - Intuition is linked to the body's interoceptive abilities, which involve the gut and heart's connection to the brain, allowing individuals to process vast amounts of information unconsciously [2][3] - The concept of "super feelers" refers to individuals who are more attuned to their intuitive responses, leading to better performance in high-stakes environments like financial trading [4][6] - Research indicates that successful traders can read physiological signals related to intuition, enhancing their decision-making capabilities [4][6] Group 2: Collective Intelligence - Collective intelligence can surpass individual capabilities when certain conditions are met, such as independence, diversity, decentralization, and aggregation of individual inputs [6][8] - The effectiveness of collective intelligence is influenced by emotional intelligence and prosocial behaviors among team members, which foster a supportive environment [9][10] - Cognitive diversity within teams is crucial for success, as it encourages different thinking styles and prevents groupthink [11][12] Group 3: Balancing Intuition and Bias - While intuition can be beneficial, it is also susceptible to unconscious biases that may lead to misjudgments, particularly under stress [7][8] - The concept of "connected thinking" is proposed as a method to balance intuition and bias, emphasizing the importance of social connections and interactions in enhancing collective intelligence [8][12]

Core Insights - The establishment of the "Shenzhen Metamaterials Industry Alliance - Drone Swarm Innovation Subcommittee" marks a significant application of metamaterials technology in the drone swarm sector, aiming to address industry pain points and promote high-quality development in the low-altitude economy [1][2] Group 1: Industry Developments - The Shenzhen Metamaterials Industry Alliance, founded in 2012, focuses on the research and development of metamaterials technology, standard formulation, and ecosystem construction to enhance China's competitiveness in the global metamaterials industry [1] - The drone swarm technology is transitioning from "technology validation" to "scale application," leveraging multi-drone autonomous decision-making and collaborative operations to overcome the limitations of individual drones [2][3] Group 2: Company Initiatives - The company has initiated a comprehensive digital platform for drone manufacturing, creating an end-to-end management system that integrates supply chain, production, and management processes, ensuring standardized and traceable digital support across the industry chain [2] - The formation of the innovation subcommittee aims to cultivate an industry ecosystem that deeply integrates metamaterials technology, digital manufacturing, and operational scenarios, fostering collaboration among industry resources and enhancing the overall strength of the low-altitude economy [3]

Core Viewpoints - The current AI models exhibit a tendency to provide answers regardless of accuracy, reflecting their nascent technological stage [2] - The rise of AI is leading to a decline in individual cognitive independence and an increased reliance on collective intelligence, effectively transferring cognitive burdens to external models [5][6] - The future may redefine life itself, with machines emerging as a new species, blurring the lines between pure humans and cyborgs [10][11] Group 1: Impact on Individual and Collective Intelligence - AI is causing a decrease in individual knowledge independence while increasing dependence on collective wisdom, a trend that has evolved from the internet and social networks to current AI models [5] - The ease of accessing vast amounts of information through AI leads to a decline in personal confidence in decision-making, as individuals struggle to determine the appropriateness of various analytical perspectives [6] - The dual nature of AI's impact should not be simplistically categorized as either "dumbing down" or "enlightening," as both effects can coexist and transform over time [6] Group 2: Future Economic and Social Structures - The future manufacturing landscape is expected to become automated and public-oriented, with production, consumption, and distribution occurring concurrently rather than sequentially [7] - Economic models will shift from being transaction-centered to focusing on individual intentions, organizing around personal interests and genuine needs [7][15] - The emergence of a "machine world" will redefine human production, organization, and consumption, leading to a potential overhaul of traditional human reproductive methods through technologies like artificial wombs [11] Group 3: Human-Machine Relationship - Discussions about human-machine relationships must adopt a long-term perspective, recognizing the need to redefine concepts of life and existence in light of advancements in biotechnology and AI [9][10] - The evolution from "human consensus" to "human-machine consensus" is crucial, requiring acceptance of machines potentially possessing free will and the need for humans to adapt to this new reality [11][12] Group 4: New Economic Logic and Cultural Integration - The transition to a new economic logic will be driven by the realization that inequality stems from mismatches rather than scarcity, leading to a focus on real-time distribution based on individual intentions [15] - The integration of Eastern and Western cultural wisdom is essential to address the limitations of current economic theories and to foster a revival of public spirit in a highly interconnected world [14][16]

Core Insights - The event is not merely about finding a "winner" but serves as a platform for technological breakthroughs and industry development through competitive rules [1][5] - The first humanoid robot competition showcases the collaboration between humans and robots, marking the beginning of a "human-robot collaboration era" [1][4] Event Overview - The humanoid robot competition took place from August 15 to 17 at the National Speed Skating Hall, featuring 127 brands and over 500 humanoid robots from 280 teams across 16 countries [4] - Unlike traditional marathons, this competition focuses on short-distance and group events, emphasizing hardware design and coordination among robots [4][9] Competition Structure - The competition includes various events such as sprinting, obstacle courses, and multi-robot soccer, assessing robots' explosive power and coordination [4][12] - Performance in these events does not solely determine the best robot, as each participant has unique strengths [5][12] Technological Focus - The competition highlights two main control methods for robots: manual remote control and fully autonomous operation, with specific events requiring complete autonomy [6] - The distinction between remote control and autonomous decision-making is emphasized, with both approaches addressing different operational needs [8] Market Implications - The event serves as a testing ground for companies to refine their technologies and potentially secure real orders, moving beyond mere demonstrations [15][17] - The integration of sports, art, and practical applications in the competition reflects a multi-dimensional approach to robot deployment in various sectors [13][16] Future Directions - The competition aims to establish a "technology-scenario" coordinate system, helping humanoid robot companies clarify their product positioning and accelerate industry maturity [17]