RT Tetsuo (@tetsuoai)🚨 Breaking AI: xAI's Grok 4 leading accessible AGI progress! New ARC-AGI SOTA via Grok 4 + open program synthesis.- jerber888 - 79.6% on V1 ($8.42/task), 29.4% on V2 ($30.40/task)- eric_pang - 77.1% on V1 ($2.56/task), 26.0% on V2 ($3.97/task)Both use Grok 4 + DreamCoder-inspired synthesis. ARC-AGI tests fluid intelligence. humans ~=98% on V1, base LLMs <10% on V2. ...

Group 1 - Dow futures fell by 0.06%, S&P 500 futures dropped by 0.14%, and Nasdaq futures decreased by 0.17% [1] - Chinese concept stocks saw pre-market gains, with Alibaba up 2.63%, Pinduoduo up 2.29%, NetEase up 1.00%, Trip.com up 1.78%, JD.com up 1.50%, and Baidu up 9.19% [1] - Morgan Stanley reiterated a "buy" rating for ASML, listing it as a preferred stock in the European tech sector with a target price of €822, suggesting that the company's worst phase may be over [1] - New Fortress Energy surged over 55% in pre-market trading after securing a $4 billion LNG supply agreement with Puerto Rico [1] - Reports indicate that Tesla may collaborate with Intel on custom chip development following a previous agreement with Samsung for custom AI chips [1] Group 2 - GlaxoSmithKline announced a commitment to invest $30 billion in the U.S. over the next five years for R&D and supply chain infrastructure, coinciding with potential import tariffs on the pharmaceutical industry [2] - Elon Musk stated that the Grok 5 model will begin training in a few weeks, with expectations that xAI could achieve AGI through this model [2] - Stifel downgraded Accenture's target price from $355 to $315 while maintaining a "buy" rating [2] Group 3 - New York gold futures briefly fell below $3,700, with a daily decline of 0.5% [3] - The Federal Reserve's FOMC is set to announce interest rate decisions and economic forecasts, with Chairman Jerome Powell scheduled to hold a press conference shortly after [3]

Core Insights - OpenAI has launched GPT-5-Codex, a refined variant of GPT-5 designed specifically for AI-assisted programming tools, which shows improved performance in coding tasks and dynamic thinking time [1][5][36] - The release of GPT-5-Codex marks a significant shift in the "coding agents" landscape, challenging the dominance of competitors like Anthropic [2][5] - OpenAI's focus on integrating research with product development has led to significant advancements in coding capabilities, with GPT-5-Codex achieving a score of 74.5% on SWE-bench, nearly matching GPT-5's score [6][36] Product Development and Features - The Codex team has worked diligently to create a multi-faceted agent capable of functioning as a software engineer, integrating various tools and interfaces such as Codex CLI and IDE extensions [6][7][18] - GPT-5-Codex exhibits enhanced endurance for complex tasks, capable of working continuously for up to seven hours, showcasing its ability to handle intricate code refactoring [8][36][37] - The model's design emphasizes a balance between speed and intelligence, allowing it to respond quickly to simple tasks while maintaining the capability for complex decision-making [36][37] Competitive Landscape - Over the past year, Anthropic has established a strong position in the coding scene, with revenues reaching $5 billion and a market cap of $183 billion, prompting OpenAI to intensify its efforts in the coding domain [5][29] - OpenAI's historical focus on programming, dating back to the original Codex release in 2021, has laid the groundwork for its current advancements and competitive strategies [5][12][14] Future Directions - The future vision for AI in programming includes a multi-agent system where numerous agents operate under human supervision, creating significant economic value [39][40] - OpenAI is committed to addressing safety and alignment issues as it develops more capable coding agents, ensuring that human operators maintain control over AI actions [39][40] - The company anticipates that advancements in AI will not only enhance coding efficiency but also unlock new capabilities in various fields, including medicine and materials science [41][42]

AI Model Performance - Grok 4 surpassed AGI benchmarks, outperforming previous results in open program synthesis [1] - Grok is currently closer to AGI compared to other AI models [1] AGI Potential - xAI's Grok 5 has a chance of achieving AGI [1] - Elon Musk believes Grok 5 could potentially reach AGI [1]

凤凰网科技讯 9月17日，马斯克在X发文表示，我现在认为xAI有机会通过Grok 5达到AGI。以前从未想 到这一点。早些时候，马斯克表示Grok 5模型将在几周后开始训练。 ...

Core Insights - Musk announced that the Grok 5 model will begin training in a few weeks [1] - Musk believes that xAI has a chance to achieve AGI through Grok 5, a notion he had not considered before [3] - Musk previously stated that he feels Grok 5 has the potential to become true AGI, a sentiment he has never felt about anything before, and he expects Grok 5 to be launched by the end of this year [5]

I now think @xAI has a chance of reaching AGI with @Grok 5. Never thought that before.X Freeze (@amXFreeze):Grok 4 just smashed the AGI benchmarks, achieving even higher score than its previous high with open program synthesisNo other model even comes close and has not passed Grok 4 previous raw performanceCurrently Grok is more closer to AGI than any other AI models https://t.co/o2PDTET44u ...

Core Insights - The report highlights that by 2035, the total computing power of society will increase by 100,000 times, with data becoming the "new fuel" driving AI development. AI storage capacity demand will grow by 500 times compared to 2025, accounting for over 70% of total storage needs. The transition from mobile internet to intelligent agent internet will see the number of connected entities expand from 9 billion to 900 billion. AI is expected to help prevent over 80% of chronic diseases, with more than 90% of Chinese households owning smart robots and AI application rates exceeding 85%, leading to a 60% increase in labor productivity [1][4][5]. Group 1: Key Technology Trends - Trend 1: AGI is expected to be the most transformative driving force in the next decade, but significant challenges must be overcome to achieve a breakthrough [4]. - Trend 2: AI agents will evolve from execution tools to decision-making partners, driving an industrial revolution [4]. - Trend 3: Development models will transform, with human-AI collaborative programming becoming mainstream, allowing humans to focus on top-level design and innovation [4]. Group 2: Interaction and Application Changes - Trend 4: Interaction methods are shifting from graphical interfaces to natural language and multi-modal interactions, enhancing user experience through voice and gesture [4]. - Trend 5: Mobile apps are transitioning from standalone functions to AI-driven service nodes, providing users with seamless experiences [4]. - Trend 6: New L4+ autonomous vehicles will become part of daily life, creating a "mobile third space" [4]. Group 3: Computing and Energy Innovations - Trend 7: The computing field will experience disruptive innovations across four core areas: computing architecture, material devices, engineering processes, and computing paradigms [5]. - Trend 8: Data will serve as the new fuel for AI, with storage capacity needs skyrocketing [5]. - Trend 9: The number of connected entities will grow significantly, marking a shift to intelligent agent internet [5]. - Trend 10: Renewable energy will become a core factor in AI development, with renewable energy generation surpassing 50% by 2035 [5].

Group 1 - Huawei held the "Exploring the Unknown, Leaping into the Future" keynote speech on September 16, officially releasing the "Intelligent World 2035" series reports, which include the "Intelligent World 2035" and "Global Digital Intelligence Index 2025" reports, forecasting key technological trends for the next decade and their impacts on various industries such as education, healthcare, finance, manufacturing, and energy [1] - The first trend indicates that AGI will be the most transformative driving force in the next decade, but several core challenges must be overcome to achieve the AGI singularity, with a necessary path towards the physical world [1] Group 2 - The second trend highlights that as large models develop, AI agents will evolve from execution tools to decision-making partners, driving an industrial revolution [2] - The third trend points to a transformation in development models, where human-AI collaborative programming becomes mainstream, allowing humans to focus on top-level design and innovative thinking while delegating tedious coding tasks to efficient AI [2] - The fourth trend notes a shift in interaction methods from graphical interfaces to natural language, evolving towards multi-modal interactions that integrate human senses, providing users with immersive experiences [2] Group 3 - The fifth trend states that mobile apps are transitioning from independent functional entities to AI-driven service nodes, where users can simply give commands, and AI agents will invoke relevant service nodes for an optimal experience [2] - The sixth trend anticipates that with breakthroughs in key technologies like world models, new L4+ autonomous vehicles will enter daily life, becoming a "mobile third space" [2] - The seventh trend predicts that by 2035, the total computing power of society will increase by 100,000 times, breaking free from traditional von Neumann architecture constraints and achieving disruptive innovations across four core areas: computing architecture, material devices, engineering processes, and computing paradigms [2] Group 4 - The eighth trend emphasizes that data will become the "new fuel" driving AI development, with AI storage capacity demands expected to grow 500 times by 2025, accounting for over 70% of total storage needs, leading to changes in storage paradigms driven by Agentic AI [2] - The ninth trend indicates that the number of connected objects in communication networks will expand from 9 billion people to 900 billion intelligent agents, marking a transition from mobile internet to intelligent agent internet [3] - The tenth trend highlights that energy will become a core constraint on the rapid development of AI, with renewable energy expected to accelerate the replacement of traditional fossil fuels, surpassing 50% of total power generation by 2035, while AI will play a central role in managing new energy systems through token management of watts for real-time energy management [3]

Core Insights - Huawei's "Intelligent World 2035" series report was officially released, highlighting key technological trends and their impacts on various industries over the next decade [2] - The report includes two major research outcomes: "Intelligent World 2035" and "Global Digital Intelligence Index 2025" [2] Trend Summaries - Trend 1: AGI is expected to be the most transformative driving force in the next decade, but significant challenges must be overcome to achieve the AGI singularity [3] - Trend 2: With the development of large models, AI agents will evolve from execution tools to decision-making partners, driving an industrial revolution [4] - Trend 3: Development models are undergoing transformation, with human-AI collaborative programming becoming mainstream, allowing humans to focus on top-level design and innovation [4] - Trend 4: Interaction methods are shifting from graphical interfaces to natural language, evolving towards multi-modal interactions that integrate human senses [4] - Trend 5: Mobile apps are transitioning from standalone functional entities to AI-driven service nodes, enhancing user experience through AI agent interactions [4] - Trend 6: Breakthroughs in key technologies will lead to the emergence of L4+ autonomous vehicles, creating a "mobile third space" in daily life [4] - Trend 7: By 2035, total computing power in society is projected to increase by 100,000 times, with disruptive innovations in computing architecture, materials, engineering processes, and paradigms [4] - Trend 8: Data will become the "new fuel" driving AI development, with AI storage capacity needs expected to grow 500 times by 2025, surpassing 70% of total storage [4] - Trend 9: The number of connected objects in communication networks will expand from 9 billion people to 900 billion intelligent agents, marking a transition from mobile internet to intelligent agent internet [5] - Trend 10: Energy will become a core constraint on AI's rapid development, with renewable energy expected to replace traditional fossil fuels, achieving over 50% of total power generation by 2035 [5]