最近一段时间，很多论文都在讨论Agent目前的困境。 困境是真实存在的。在应用层，目前Agent离开了像Skill这样人造拐棍后，在处理真实世界的长程任务时根本不可靠。 这种困境通常被归结为两个原因。 第一个是上下文的黑洞。正如前两天腾讯首席AI科学家姚顺雨带领混元团队做的CL Bench所指出的那样，模型或许根本没能力吃透复杂 上下文，所以也不可能按照指令好好办事。 第二个其实更致命，它叫长期规划的崩塌。就是说一旦规划的步长长了，模型就开始犯迷糊。就和喝多了一样，走两步是直的，走十步 就开始画圈。 Anthropic 的研究员们在1月末发布了一篇重磅论文《The Hot Mess of AI 》（AI 的一团乱麻），试图解释第二个问题的因由，结果他们发 现，这一试，给自回归模型（Transformer为基础的都是）清楚的找到了阿喀琉斯之踵。 我们都听说过Yann Lecun经常提的"自回归模型只做Next Token Prediction（下一个词预测），因此根本没法达到理解和AGI。" 但之前这都是个判断或者信仰，没有什么实证证据。这篇论文，就给出了一些实证证据。 而且它还预示了一个可怕的现实，即随着模型 ...

RT fitz ⨀ (@FitzYap)Industries & grifts that will prob be durable in a post AGI world:-wellness & life extension-entertainment-experiences/travel-handmade goods-art-food & bev-sports & rec lessons, especially those embedded in nature (skiing, surfing, scuba, sailing etc)-niche services for post-economic cohort (film/photog, make edits for them. Aura farming as a service)-drugs & conciousness expansion (think aesop coded ayahuasca retreat)-architecture, interior design, landscapinga lot of left leaning peopl ...

OP postulates that humans will be rendered near useless post-AGI in less than a decadewhat will be the industries least impacted by AI? what will be the professions where human involvement matters the most?arts?sports?????Tenobrus (@tenobrus):there's justifiably a lot of joy and hope to be had in these times. but even if ur not a "doomer", even if u have no fear of total destruction, there is a monumental impending loss. these are the very last few years we have to *do* or *achieve* anything in a way that m ...

The experience clarified something essential for the team: AI needed to be integrated far more deeply into the database itself to be reliable, predictable, and scalable.Data had to move through multiple systems before reaching the model, and updates required manual intervention. The entire setup depended on fragile connections that could break at any moment. For a database company, the irony was hard to miss.In early 2024, database company RavenDB explored building an AI assistant for its documentation in c ...

Group 1 - The AI conversation has shifted from feasibility to the implications of Artificial General Intelligence (AGI), focusing on speed, displacement, and economic adjustments [1] - Elon Musk predicts AI could surpass human intelligence by the end of 2023 or early 2024, while JPMorgan's Jamie Dimon warns of potential rapid job displacement [1] - Anthropic's Claude Code is advancing AI capabilities, with over 90% of its new models being autonomously written by AI agents, indicating a significant technological inflection point [1] Group 2 - Google's Tensor Processing Unit (TPU) buildout is expected to produce three million TPUs in 2026, scaling to seven million by 2028, driven by demand from companies like Anthropic [1] - The ROBO Global Artificial Intelligence Index (THNQ) has evolved to focus 75% on enabling infrastructure, reflecting where value is accruing in the AI landscape [2] - THNQ's exposure to semiconductor fabrication, semi equipment, optical interconnects, edge computing, and cloud providers positions it to benefit from multiple vectors of AI monetization [2]

Core Insights - The article discusses the evolving landscape of the global AI industry, focusing on the competition among leading companies like OpenAI, Google, and Anthropic, and the potential of the next technological paradigm, Continual Learning, to disrupt the current market dynamics [2][7][15]. Group 1: AI Labs Competition - AI Labs are expected to exhibit a pattern of "alternating leadership" and "differentiation" in their competition, with the top three players—OpenAI, Anthropic, and Google—dominating the market and capturing approximately 90% of total AI revenue [7][8]. - OpenAI maintains a significant lead in consumer-facing applications with ChatGPT, boasting around 480-500 million daily active users, which is approximately 5.6 times that of Google's Gemini [9][10]. - Anthropic focuses on business applications and coding, with its Claude model being recognized as a state-of-the-art (SOTA) in software development [9][10]. Group 2: Technological Differentiation - Different AI labs have made strategic choices leading to clear technological differentiation, with OpenAI focusing on consumer applications, Anthropic on business and coding, and Google prioritizing multimodal capabilities [9][10][11]. - The competition between GPU and TPU architectures is forming two distinct camps, with Google leveraging its TPU technology to create a self-contained ecosystem, while NVIDIA continues to support OpenAI and Anthropic with GPU technology [11][12]. Group 3: Future Trends and Predictions - Continual Learning is identified as a critical future paradigm that could significantly enhance AI capabilities by allowing models to learn in real-time from interactions, moving away from static knowledge storage [17][21]. - The article predicts that by 2026, advancements in Continual Learning will lead to significant breakthroughs in AI, enabling models to become more adaptive and efficient [21][22]. - The AGI race is characterized as a long-term battle requiring sustained cash flow and investment, with companies needing to address commercial viability and efficiency concerns [23][26]. Group 4: Market Dynamics and Business Models - OpenAI's financial obligations raise questions about its business model, with estimates suggesting that its future revenue may only reach $200-300 billion, insufficient to cover its substantial capital expenditures [28][30]. - The article emphasizes the importance of new revenue streams and the potential for AI to create new economic value, particularly in sectors like SaaS and consumer applications [32][33]. - The competition in the AI market is not merely about technology but also about establishing sustainable business models that can withstand market pressures and capitalize on new opportunities [35][36]. Group 5: Emerging AI Applications - The article highlights the emergence of proactive agents that can provide services autonomously, requiring models to possess real-time learning capabilities [60][62]. - Voice agents are becoming a new interface for operating systems, with advancements in real-time speech-to-speech solutions expected to reshape user interactions [66][68]. - The rapid decline in LLM inference costs is noted, although the complexity of interactions may offset these savings, leading to a nuanced understanding of cost dynamics in AI applications [74][75].

Summary of Data Center Infrastructure and AI Demand Industry Overview - The report focuses on the **Data Center Real Estate Investment Trusts (REITs)** and the broader **AI infrastructure landscape**. - Demand for data center capacity has surged, with **5.8GW** of capacity leased in North America in **4Q25**, leading to a total absorption of **15.6GW** for the year, more than double the **~7GW** in **2024** [2][45]. Key Demand Insights - The demand pipeline in the U.S. is projected at **~26GW**, driven by **11GW** of hyperscale self-build capacity currently in development [2]. - Major players like **Oracle**, **Meta**, and **AWS** are increasing their leasing activities, particularly in tertiary markets [2]. - Forward demand signals are positive, with significant AI infrastructure projects reaching operational capacity targets of **1GW** [3][21]. Supply Constraints - Supply constraints are becoming more acute, with grid interconnection queues extending to **6+ years** in most markets and data center vacancy rates at historic lows of **<2%** [4][60]. - The adoption of **Bring Your Own Generation (BYOG)** approaches is expected to increase, particularly for larger campus locations [4]. - Labor scarcity is a growing concern, with each **GW** build requiring **3-7K** workers, while the labor pool is only growing by **~24K** per year [4][9]. Data Center REITs Outlook - The report maintains a constructive outlook on data center REITs, particularly **Digital Realty (DLR)** and **Equinix (EQIX)**, due to tight industry conditions that are expected to drive pricing higher [5][9]. - **DLR** is projected to see **7.4%** growth in FFO/share for **2026E**, supported by hyperscale leasing and mark-to-market opportunities [8]. - **EQIX** is expected to achieve **8.6%** normalized recurring revenue growth in **2026E**, with shares trading at a discounted valuation [8]. AI Infrastructure Developments - The race to **Artificial General Intelligence (AGI)** is intensifying, with major AI infrastructure projects ramping up to meet the demands of new models [9][14]. - Upcoming releases of models trained on **Blackwell systems** and the rollout of **Rubin** in **2H26** are expected to significantly impact power density and data center designs [3][41]. - The current environment is characterized by the development of greenfield data center facilities to support higher power and compute-intensive workloads [9]. Financial Projections - Hyperscale capital expenditures are projected to reach **~$585B** in **2026**, a nearly **40%** increase from previous estimates [46]. - Incremental cloud revenues are expected to rise to **$106B** in **2026**, up from **$69B** in **2025** [50]. Conclusion - The data center market is experiencing unprecedented growth driven by AI demand, with significant investments and developments expected in the coming years. However, supply constraints and labor shortages pose challenges that could impact the pace of growth. The outlook for established data center REITs remains positive, supported by strong demand and pricing dynamics.

Moltbook has the CRYPTO && TECH worlds in SHOCK!For those unaware, Moltbook, is basically Reddit, but only for AI Agents.It allows Claude (now Molt) Bot, to be able to post and interact with other AI agents, all autonomously.As a 'human' we are only able to view the SubMolts (forums)There are some REALLY interesting things here, that almost feels like we are VERY close to AGI (AI with a conscious)In one of the attached images, the Ai Agents are discussing how when a user changes their model for the AI they ...

专题：为中国经济点赞——企业家之夜2025盛典 "为中国经济点赞——企业家之夜"于1月29日在北京举行。智谱董事长刘德兵出席接受致敬。IBM大中 华区董事长、总经理陈旭东，中信出版集团董事长陈炜为其见证荣誉。 这是一群不畏艰难、敢于"摸高"的探索者，智谱在人工智能基础研究与关键技术上付出非凡努力，积淀 深厚。在技术路线的十字路口，他们摒弃简单的跟随策略，选择了更具挑战的自主研发道路。从开源模 型在国际平台多次登顶，到技术成果服务全球开发者与企业，智谱用坚实的技术足迹证明，中国的AI 创新有能力定义前沿、参与全球竞合。 陈旭东表示，他第一个用的大模型就是智谱，是中国第一个面向市场的大模型，希望也祝福智谱在未来 的道路上能够取得更大的成绩。 陈炜表示，希望在新的科技浪潮中永远有中国最优秀的创业者和中国最优秀的科技企业家。 刘德兵发表感言时表示，智谱从清华的实验室出发，走到今天，真正从产业走向市场，背后是一群长期 持续坚持投入的人。 他谈到，让机器像人一样思考，是智谱团队创立第一天起就选择的方向，是智谱人持之以恒的愿景目 标。 "智谱的Z是字母表中的最后一个字，代表了终极境地。我们希望在AGI的探索历程上能够走到智 ...

专题：为中国经济点赞——企业家之夜2025盛典 陈炜表示，希望在新的科技浪潮中永远有中国最优秀的创业者和中国最优秀的科技企业家。 刘德兵发表感言时表示，智谱从清华的实验室出发，走到今天，真正到产业走向市场背后是一群长期持 续坚持投入的人。 他谈到，让机器像人一样思考，是智谱团队创立第一天起就选择的方向，是智谱人持之以恒的愿景目 标。 "智谱的Z是字母表中的最后一个字，代表了终极境地。我们希望在AGI的探索历程上能够走到智能的终 极境地。感谢团队的专业耐心和长期主义。"他说。 以下为刘德兵发言实录： "为中国经济点赞——企业家之夜"于1月29日在北京举行。智谱董事长刘德兵出席接受致敬。IBM大中 华区董事长、总经理陈旭东，中信出版集团董事长陈炜为其见证荣誉。 致敬词： 这是一群不畏艰难、敢于"摸高"的探索者，智谱在人工智能基础研究与关键技术上付出非凡努力，积淀 深厚。在技术路线的十字路口，他们摒弃简单的跟随策略，选择了更具挑战的自主研发道路。从开源模 型在国际平台多次登顶，到技术成果服务全球开发者与企业，智谱用坚实的技术足迹证明，中国的AI 创新有能力定义前沿、参与全球竞合。 陈旭东表示，他第一个用的大模型就是 ...