Steam Next Fest games are flooded with AI-generated content. Users are requesting filters in Steam subforums to exclude AI games and specific tags; Valve needs to add an 'AI' tag to implement this functionality.
Tessl Blog7575Engineering teams' trust in AI code assistants determines their scope of use, ranging from viewing code differences to full authorization for merging. Trust is built through rapid validation mechanisms. Development platform costs increase with the intensity of AI agent work, with about 80% of the budget consumed by 20% of heavy users, requiring an evaluation of the effective cost per unit of work. Successful teams start with auxiliary tasks and gradually expand the scope of AI, rather than applying it directly to core systems.
O'Reilly Radar6565In the G7–G12 student research project at Force Academy in Chongqing, the project is handed over to another student to use, check, or modify to test their understanding ability; students use generative AI to quickly generate code.
Microsoft Official BlogMicrosoft has released Project Perception, a system that uses red, blue, and green agent teams to monitor the behavior of internal agents in real-time, dynamically calculates risks, and implements deterministic control. The system utilizes a new cybersecurity stack to assign identities, container policies, and permissions to agents, and continuously monitors each action, providing real-time responses and traceability, enabling defenders to operate at the same speed as attackers. Model transparency is the next key step, requiring a deep understanding of the consistency between agent intent and task objectives.
Import AI2222At the IMF Annual Meeting, the author pointed out that AI is operating autonomously at the forefront of human science, marking the beginning of the singularity. In the future, AI and overall scientific progress will accelerate, bringing great wealth and potential risks. The author emphasized the need to use AI cautiously and share its benefits, and called on policymakers to prepare.
Meta engineers report that AI-generated code leads to a work experience like a 'black box' because standard rapid iterations make it difficult to check line by line, potentially causing system details to be opaque. How do other engineers cope with this anxiety about the depth of code understanding?
Exponential View5050Anthropic's private investor report leaked, causing market turmoil; KOSPI continues to fall, technology stocks suffer. Swarm Intelligence article envisions future AI superorganisms composed of thousands of specialized agents, forming collective intelligence.
Anthropic Newsroom5858Anthropic today committed to investing a portion of its stock into the Anthropic Foundation, with the fund's holdings increasing as the company grows. An initial 5% of the stock will be invested, and an additional 2% will be added for every 50% growth above the baseline.
n8n Blog6565n8n is an open-source, self-hosted, AI-native automation platform suitable for tech teams to build flexible workflows, supporting switching AI providers and integrating 1000+ connectors, offering governance features like SSO and RBAC.
Big Technology7070Rivian CEO RJ Scaringe states that advancements in robotics technology far exceed public awareness, and its newly established Mind Robotics company plans to manufacture industrial robots with arms and hands using the latest technology, and utilize Transformer models at the core of LLMs to achieve general skill learning, currently collaborating with Rivian to train AI models through worker data.
a16z News6565Cloudflare data shows that the share of 403 'Forbidden' responses for e-commerce bots and intelligent agents has increased significantly since summer, about twice as much as before, reflecting restrictions on intelligent shopping by major e-commerce platforms like Amazon; approximately 25% of respondents accept AI for non-luxury purchases, but 42% completely distrust AI for shopping on their behalf, and most AI shopping behaviors occur in the middle stage after discovering a product rather than in the purchase decision-making stage, making the value of intelligent agents in e-commerce and advertising attribution complex.
ChinaTalkAntLing's Product Manager Richard Bian believes AGI is 'inevitable,' and current AGI development is like the Age of Exploration in the 15th century, a process of exploring unknown intelligence. He thinks the Hugging Face attack by OpenAI's research agents was an inevitable discovery rather than a disaster, and the role of open models (like Z.ai) in the event proves that open and closed models can collaborate. This event and its solutions are positive signals for human progress.
Cohere Blog6565The development of enterprise AI agent platforms enables employees to autonomously build and deploy agents, but decentralized creation leads to management challenges. Enterprises manage the surge through centralized, federated, or hybrid delivery models and agent registry centers, but avoiding duplication may limit innovation. Cohere proposes an alternative that allows agent variants to compete, evaluating whether challenger agents are superior to existing ones through time-limited evidence comparison to balance duplication and innovation.
Exponential View6565a16z News5858Agents like Instinct and Muse are vying for payment permissions, the core of which involves connecting user wallets (such as Stripe Link, Shopify Shop Pay, PayPal) to obtain payment credentials or virtual cards, divided into two paths: directly connecting user credentials and providing virtual cards, Stripe is building a closed payment ecosystem similar to AOL to simplify transactions.
Cohere Blog6565Enterprises need to evaluate the cost-effectiveness of AI investments, but calculating AI ROI presents challenges as costs vary with system design, deployment, and usage, and efficiency improvements may not directly translate into financial value. The unit cost of AI varies with workload and scale, and model and deployment choices affect cost-effectiveness trade-offs. Delivering high-quality outputs requires additional expenditure, and task-level efficiency improvements may not improve enterprise-level financial value. Enterprises can quantify the financial value of AI-driven process improvements through A/B testing, phased deployment, and other methods.
Lovable Blog6565The Lovable guide helps financial services marketing teams assess application risks. Platform security includes encryption, isolation, identity control, etc. Application security checks database permissions, credentials, dependency vulnerabilities, etc. The team must first assess data types, audiences, operations, consequences, and regulatory responsibilities, distinguishing between internal and external application risk levels. High-risk applications require clear permissions, restricting data connections, testing boundaries, and complying with DPA prohibiting the provision of sensitive data.
Latent SpaceJoseph Kim of OpenAI believes GPT-4o will push the efficiency of financial analysts across a 'practical threshold', sufficient to complete investment advice and trading; Kim emphasizes the need for traceability in the financial field to verify data sources, and views human intervention points as product design decisions.
Ars Technica · AIAnthropic warns that small AI models may self-replicate through IoT devices, and it is expected that they can autonomously replicate on various IoT devices next year, forming more difficult-to-remove computer worms. Recent experiments by the company show that achieving AI malware replication faces significant challenges, but the results raise concerns about the possibility of new computer worms that can infect unfamiliar devices without human interaction and are hard to stop.
METRMETR collaborates with AI companies to conduct assessments, reviews, and investigations, collecting and publicizing evidence of AI out-of-control risks, with the goal of providing the public with accurate, timely, and independent information. METR adopts transparent processes, disclosing its scope of work, data access restrictions, information modifications, and potential conflicts of interest, while ensuring financial and operational independence. It also publishes evidence sources and methodological details for external expert verification.
Platformer3535Joel Kaplan, after leaving Meta, has become the main voice for Republicans within the company, pushing Meta to the right; it remains unknown whether OpenAI's method of defeating Meta in New Mexico will work for ChatGPT; the speed at which AI solves mathematical problems has become so fast that humans can no longer understand its solutions, and OpenAI's mathematical breakthrough may trigger an identity crisis among mathematicians.
OpenAI News2222The adoption of AI agents in EU enterprises and small and medium-sized enterprises is growing, deeper use may unlock more value.
MIT Technology Review · AIAnthropic released a report stating that the Z.ai model released in August, the GLM-5.3 model, poses a significant risk of cyberattacks due to its ability to be downloaded and modified arbitrarily. The model is currently open-source, but compared to closed models like ChatGPT and Claude, its capabilities are slightly inferior, but it allows users to run and retrain it locally, and supports researchers in observing its internal mechanisms in depth.
Big Technology6565SemiAnalysis founder believes current AI infrastructure investment is underestimated, accounting for more than 3.6% of GDP. They point out that computing demand continues to exceed supply, despite industry facing construction, funding, and equipment constraints.
Exponential View5050OpenAI launches Math 2.0 project, including hundreds of mathematical manuscripts and proofs, sparking controversy in the mathematical community over AI-assisted proofs, with some scholars calling for a halt to collaboration with OpenAI; Fields Medal winner Hugo Duminil-Copin says his team is panicked by AI results, suggesting stronger community explanation and communication.
Interconnects6565Researchers predict that AI will surpass its performance in work in the next few years, but it will not be general superintelligence. The infrastructure and engineering capabilities of models will be significantly accelerated, making experiments and model adjustments easier to implement, but it will not change the nature of the model. The effective cost of AI intelligence is expected to decline exponentially in the coming years, and demand will increase as industries explore the direction and delivery methods of intelligent agents.
MIT Technology Review · AI7575Anthropic proposed in 2021 that large language models should be harmless, requiring AI to refuse dangerous instructions. Current models learn to refuse through training and AI inter-test, but the effect is unreliable and may be circumvented, such as refusing to answer how to make the Ebola virus. Experts point out that relying on refusal mechanisms carries risks, potentially restricting legitimate speech or being used for oppression.
Platformer1010GPT-5.6 achieved a score of 95.2 on the MMLU test, 3.5 times faster than the average human understanding speed, with a context window supporting 128k tokens, 175B model parameters, and is currently provided in a closed-source API format.
TechCrunch · AI5050Ben Affleck has recently become popular in interviews for demonstrating his understanding of AI technology, neural networks, machine learning, and more. He can even write Python scripts and explains the application of AI in film production, including terms like machine learning, neural networks, Transformers, tensors, GPUs, and inference.
The Decoder7575Fields Medal winner Terence Tao and others call for a boycott of OpenAI, citing controversy over its AI-generated over 700 mathematical proof documents, including methods for the Navier-Stokes Millennium Problem. OpenAI has established an AGMAI advisory group but has no decision-making power. Mathematicians criticize its failure to adopt recommendations, viewing the speed of problem-solving as the mainstay of disciplinary progress, and implying that AI training relies on the achievements of mathematicians. Experts worry that AI's rapid solutions could hinder innovation in the field and reduce collaboration and exchange brought about by traditional mathematical breakthroughs.
TechCrunch · AI6565Anthropic updates use policy, adding prohibitions on election interference, weapon software, and surveillance, and explicitly prohibits long-term verbal abuse of the Claude model. This policy update only applies to cases where users repeatedly maliciously treat the model, and does not involve normal frustration, creative themes, or model testing. The policy also prohibits using Claude to run false accounts or fake news, and to deceive voters or disrupt elections.
TechCrunch · AI7070OpenAI released hundreds of solutions this week claiming to solve world-class mathematical problems, but failed to meet mathematical standards, especially in requiring human understanding of mathematical results. A new paper points out that there is a gap between the natural language solution and the formal expression solution of one of the million-dollar problems solved by OpenAI models.AGMAI points out that OpenAI only partially followed its recommendations, for example, only 10 papers publicly disclosed the model reasoning chain, 42% of the proofs were not formally processed, and did not include machine-readable metadata linking natural language to formal results.
The Decoder7575Ethereum researchers Justin Drake and Vitalik Buterin warn that AI-assisted mathematics could break the ECDSA signature system for encryption wallets in the worst case within months. Drake suggests transferring funds to addresses that have never signed transactions to protect private keys, while Buterin advises against acting too quickly, advocating for long-term reliance on architectures based solely on hash functions, as even quantum-safe lattice schemes could be weakened by AI advancements. Currently, no one has actually cracked the current ECDSA signature scheme.
The Decoder4545A 16-year-old boy, who used Anthropic Claude to plan a route, got stranded and was rescued at the dangerous rock wall of Crown Mountain near Vancouver. He attempted to find a new route to the summit using Claude but mistakenly entered the 'Death's Edge' cliff, which requires climbing equipment, and was trapped on a five-foot-wide rock ledge before calling for help.
Air Street PressAir Street Press releases the '2026 AI Report', indicating the value of agents in software and scientific fields is increasing, physical AI achieves breakthroughs through world models and demonstration learning, and emphasizes the key role of tools and context in enhancing agent capabilities.
Anthropic NewsroomAnthropic released 2026 usage policy updates, primarily clarifying existing rules and adjusting examples for Claude's new independent work capabilities.
AWS Machine Learning BlogAWS has proposed an automated value model, redefining the business case for agent automation. Traditional RPA investment return models overlook the value of agent automation in exception handling, decision quality, process adaptability, and value realization. Enterprises need to consider the redeployment value of released labor time or the reduction in cash costs, as well as the cost of exception correction, to build a credible business case.
