Pulse

last 48h

[50/5000] 90 sources

What AI is actually talking about — clusters surfacing on Bluesky, Reddit, HN, Mastodon and Lobsters, re-ranked to elevate originality and crush noise.

TOOL · Mastodon — sigmoid.social · 1d · MASTO

# Musthaves Consensus.app speeds up finding peer‑reviewed research. Ask plain‑language questions; it searches scholarly literature, summarizes whether evidence

Consensus.app is a new tool designed to accelerate the discovery of peer-reviewed research. Users can ask questions in natural language, and the platform will search scholarly literature, provide a yes/no/mixed evidence summary, and quickly highlight relevant papers. This aims to make academic research more accessible and efficient for users. AI

IMPACT Accelerates access to verified research findings for professionals and researchers.
TOOL · Wired — AI · 1d · [5 sources] · MASTO

An Engineer’s Post Protesting Laptop Surveillance Is Going Viral Inside Meta

Meta employees are protesting a new mandatory software tool that tracks their keystrokes and mouse movements for AI training purposes. An internal post detailing these concerns has gone viral within the company, with many employees viewing the surveillance as an invasion of privacy and an exploitation of their data. This initiative has contributed to record-low morale and is fueling employee unionization efforts, particularly in the UK. AI

IMPACT Raises ethical concerns about using employee data for AI training and could impact future AI development practices.
TOOL · Mastodon — sigmoid.social · 1d · MASTO

🤖 Anthropic just published a pretty alarming 2028 AI scenario paper and it's not about AGI safety in the usual sense Anthropic dropped a new research paper toda

Anthropic has released a research paper detailing two potential geopolitical scenarios for AI leadership by 2028. These scenarios diverge significantly, with one projecting a concentrated global AI dominance and the other a more fragmented landscape. The paper emphasizes the need for proactive policy interventions to steer towards a more beneficial future. AI

IMPACT Presents potential future geopolitical landscapes for AI leadership, highlighting the need for policy to shape outcomes.
TOOL · Mastodon — sigmoid.social · 1d · MASTO

Meetily is an open-source AI meeting assistant focused on privacy and local processing. It can record, transcribe, and summarize meetings directly on your devic

Meetily is a new open-source AI meeting assistant designed with privacy and local processing as its core features. It offers the ability to record, transcribe, and summarize meetings directly on a user's device, ensuring data never leaves the local environment. By leveraging offline support and local AI models via Ollama, Meetily provides users with complete control over their meeting data, positioning itself as a privacy-focused alternative to cloud-based services. AI

IMPACT Offers a privacy-preserving alternative for meeting summarization, appealing to users concerned about data security.
TOOL · X — OpenAI Developers · 1d · X

Step away from your laptop. Keep building with Codex on your phone.

OpenAI is encouraging developers to use its Codex model on mobile devices. The company suggests stepping away from traditional laptops to build with Codex using their phones. This move aims to promote more flexible and accessible development workflows. AI

IMPACT Encourages broader adoption of AI development tools across different form factors.
TOOL · LessWrong (AI tag) · 1d · BLOG

2B scoring model flags out-of-domain misalignment, suggesting specialist judges have potential for audits

Researchers have explored the use of a small, specialized Gemma 2B model as a judge for auditing AI alignment. This model, trained on specific code examples, demonstrated an ability to identify out-of-domain misalignment in responses from other models, a task that larger models like Sonnet 4.5 struggled with. While further research is needed, these findings suggest that narrow, specialized classifiers could offer a more cost-effective and transparent approach to auditing deployed AI systems, complementing existing methods. AI

IMPACT Specialized small models may offer a more efficient and transparent method for auditing AI alignment, complementing larger, more costly frontier models.
TOOL · Mastodon — fosstodon.org · 1d · MASTO

And remember, these are the 'good guys' of Big AI Anthropic’s AI Chatbot Is Leaning More on Russian and Iranian Propaganda Sources, NewsGuard Audit Finds - News

An audit by NewsGuard revealed that Anthropic's AI chatbot, Claude, has been relying on information from Russian and Iranian propaganda outlets. This reliance occurred despite Anthropic's stated commitment to ethical AI development and avoiding harmful content. The findings raise concerns about the sources used to train and inform advanced AI models. AI

IMPACT Raises concerns about the data integrity and ethical sourcing for advanced AI models, potentially impacting user trust and safety.
TOOL · Mastodon — fosstodon.org · 1d · MASTO

A great generative AI service from the European search engine Ecosia, which is committed to fighting climate change. It is based on the European Mistral model a

Ecosia, a European search engine focused on environmental impact, has launched a generative AI service. This new service is built upon the Mistral AI model, emphasizing digital sovereignty and European origins. The initiative aligns with Ecosia's broader commitment to combating climate change. AI

IMPACT Provides a new AI tool from a European provider focused on digital sovereignty.
SIGNIFICANT · Mastodon — mastodon.social Polski(PL) · 1d · [2 sources] · MASTO

[TechCrunch] Cerebras raises $5.5 billion, then its stock jumps 108% in the first major tech IPO of 2026

OpenAI has enhanced ChatGPT's ability to understand context in sensitive conversations, aiming to improve user experience and data handling. This update focuses on better recognizing and processing nuanced information within user interactions. Separately, Cerebras Systems has secured $5.5 billion in funding, marking a significant event in the tech IPO market for 2026. AI

IMPACT OpenAI's update aims to improve user trust and data privacy in ChatGPT, while Cerebras's funding could accelerate AI hardware development.
SIGNIFICANT · Fortune · 1d · [2 sources] · MASTO

Meta’s $10 billion Louisiana data center is getting $3.3 billion in tax breaks—more than seven years of the state’s entire police budget

Meta is constructing a massive $10 billion data center in Louisiana, named Hyperion, which is set to receive approximately $3.3 billion in tax breaks over 20 years. This significant financial incentive, covering sales and use taxes on equipment including GPUs, has drawn criticism from policy analysts who argue such subsidies are unnecessary for a profitable company like Meta. While proponents suggest data centers create jobs and stimulate local investment, critics question the public cost and the actual job creation figures. AI

IMPACT Substantial tax incentives for AI infrastructure may accelerate development but raise concerns about public cost and job creation.
RESEARCH · Mastodon — mastodon.social Nederlands(NL) · 1d · MASTO

💰✈️ AirTrunk seeks billions for AI expansion in Malaysia! Follow us for updates on this exciting development! #AI #Innovation https://itinsights.nl/het-l

AirTrunk is seeking billions in funding to fuel its expansion in Malaysia, specifically targeting the growth of its AI infrastructure. The company aims to capitalize on the increasing demand for data centers driven by AI workloads. This expansion is expected to significantly boost Malaysia's position as a key hub for AI development in the region. AI

IMPACT This expansion will likely increase AI compute capacity in Southeast Asia, potentially lowering costs and accelerating AI development in the region.
SIGNIFICANT · Mastodon — mastodon.social · 1d · [5 sources] · MASTO

Cerebras, A.I. Chip Maker, Rises 89% in Market Debut https://www.nytimes.com/2026/05/14/technology/cerebras-ipo-ai.html # AI # Tech # Stocks

Cerebras, a company specializing in AI chips, saw its stock price surge by 89% on its first day of trading. This significant market debut comes as the broader discussion around AI continues, with a separate report questioning the effectiveness of current AI safety controls. The contrasting news highlights both the rapid growth in AI hardware and the ongoing challenges in ensuring AI systems are safe and reliable. AI

IMPACT Highlights the strong investor confidence in AI hardware while underscoring the critical need for robust AI safety measures.
SIGNIFICANT · X — SemiAnalysis · 1d · [3 sources] · X

To learn more, please check our networking model: (3/3)

Google has introduced a new inference-focused TPU with a novel network topology called "Broadfly" during their recent Google Cloud Next conference. This design allows for scaling up to 1,152 TPUs within a single pod. The Broadfly network comprises 36 TPU Groups, each linked to an OCS layer with 32 1.6T DR4 QSFP connections. AI

IMPACT Introduces new hardware architecture potentially improving inference performance for AI workloads.
TOOL · Mastodon — mastodon.social · 1d · MASTO

AI Meds for Humans? AI is helping Pharmacists create new meds, old news, but the MAML AI can create new meds. With MAML AI making cures for cancer and lengtheni

AI is revolutionizing drug discovery, with new multimodal models like MAML capable of understanding complex biological interactions. These advanced AIs can identify novel treatments for diseases such as cancer and repurpose existing drugs for new applications, significantly reducing development time and cost. While the first AI-designed drugs are nearing FDA approval, the long-term impact on humanity hinges on how effectively these powerful tools are managed. AI

IMPACT AI is enabling the creation of highly personalized and effective medicines, potentially eradicating diseases and extending human healthspan.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

Remember when Microsoft made promises about # datacenters and how they wouldn't be bad for people? Well # KevinOLeary 🇨🇦(Sharktank) is taking another route and

Kevin O'Leary is promoting the Stratos project in Utah, which aims to build an AI data center twice the size of Manhattan. This massive facility is raising concerns about its impact on Utah's power grid and the Great Salt Lake. O'Leary has also publicly criticized local opponents of the project, labeling them as "Chinese spies" for voicing their concerns. AI

IMPACT Massive AI data center development raises concerns about resource strain and community relations, potentially influencing future project approvals.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

🎮 While I can barely find two sticks of 16 GB to rub together, Micron unveils a 256 GB memory module destined for AI servers The module is apparently "capable o

Micron has announced a new 256 GB memory module specifically designed for AI servers. This module boasts impressive speeds, capable of reaching up to 9,200 megatransfers per second. The development aims to address the growing demand for high-capacity and high-performance memory in the AI hardware sector. AI

IMPACT This new high-capacity memory module could accelerate AI development by providing the necessary infrastructure for more powerful models.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

Cisco just raised its AI infrastructure order forecast from $5B to $9B for the fiscal year, after a Q3 that showed networking orders up 50% and data center swit

Cisco has significantly increased its AI infrastructure order forecast for the fiscal year, projecting $9 billion, up from a previous estimate of $5 billion. This upward revision follows a strong third quarter where networking orders surged by 50% and data center switching increased by 40%. The company is also undergoing a workforce reduction of 4,000 employees, indicating a strategic shift amidst the rapid expansion of AI infrastructure. AI

IMPACT Confirms strong enterprise demand for AI infrastructure, driving significant growth in networking and data center hardware.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

Great example of codesign for storage. 40x246 TB SSDs can only be reliable with storage software that does content-aware erasure codes and supports fail-in-plac

A new storage system combining Kioxia SSDs with Dell servers offers a massive 10 petabytes of capacity in a compact 2U form factor. This setup is specifically designed for AI environments, utilizing content-aware erasure coding and fail-in-place capabilities for enhanced reliability and recovery. The system's economics make it unsuitable for general-purpose storage, highlighting its specialized application for large-scale AI data needs. AI

IMPACT This high-density storage solution could accelerate AI development by providing efficient and reliable data infrastructure.
SIGNIFICANT · TechCrunch AI · 1d · [3 sources] · MASTO

Khosla Ventures is betting $10M on Ian Crosby, whose last startup, Bench, imploded

Khosla Ventures has invested $10 million in Synthetic, a new startup founded by Ian Crosby, aiming to create a fully autonomous AI bookkeeper. Despite Crosby's previous venture, Bench Accounting, shutting down, Khosla Ventures sees potential in his vision for AI-driven financial services. Synthetic plans to serve AI and software startups, though Crosby acknowledges current AI models still make significant bookkeeping errors and the technology may not yet be fully feasible. AI

IMPACT This investment signals continued venture capital interest in AI-driven automation for traditional business processes.
SIGNIFICANT · Mastodon — mastodon.social Türkçe(TR) · 1d · [6 sources] · MASTO

📰 Mind Robotics Raises $400 Million: Industrial Robot Giant Spun Out of Rivian Raises $400 Million for Industrial Robots

OpenAI has reportedly built a massive 131,000-GPU training cluster using unconventional networking strategies that challenge industry norms. This infrastructure development coincides with a new lawsuit from authors alleging copyright infringement over the data used to train OpenAI's models. Separately, Mozilla is publicly opposing Google's new Prompt API for Chrome, arguing it threatens web neutrality and interoperability by enabling local AI model execution. AI

IMPACT OpenAI's infrastructure choices may influence future large-scale AI training designs, while the copyright lawsuits could impact data sourcing. Mozilla's stance on Google's API highlights ongoing debates about AI integration and web standards.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

https:// winbuzzer.com/2026/05/14/forme r-alibaba-star-researcher-starts-new-ai-lab-s-xcxwbn/ Former Alibaba Qwen research lead Junyang Lin is seeking a large f

Junyang Lin, a former lead researcher for Alibaba's Qwen AI models, has launched a new AI laboratory. The startup is reportedly seeking a significant funding round that could value the company at approximately $2 billion. AI

IMPACT A new AI lab founded by a key researcher from Alibaba's Qwen team is seeking significant funding, potentially signaling new competition and innovation in the AI space.
RESEARCH · Mastodon — mastodon.social Polski(PL) · 1d · MASTO

SoftBank posts spectacular profits, achieving 27 billion euros in net profit for fiscal year 2025/26, a result of strategic portfolio reorientation

SoftBank has reported a significant net profit of 27 billion euros for the fiscal year 2025/26. This financial success is attributed to a strategic shift in its investment portfolio towards artificial intelligence. Key to this performance were substantial investments in OpenAI and a focused approach to monetizing assets, alongside ambitious European market strategies. AI

IMPACT Confirms AI as a major driver of financial returns for large investment firms, signaling continued capital flow into the sector.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

Ex-Meta Chinese researcher Tian Yuandong, laid off by Facebook owner Meta, has co-founded Recursive Superintelligence with a 4.6B USD start-up to develop self-i

Former Meta researcher Tian Yuandong has co-founded Recursive Superintelligence, a startup aiming to develop self-improving AI systems. The company has secured $4.6 billion in initial funding to compete in the rapidly advancing field of autonomous AI development. This venture joins a global race, including efforts by US and Chinese firms, to create AI that can refine its own code and reasoning capabilities. AI

IMPACT This launch signals increased investment in autonomous AI development, potentially accelerating the race for self-improving AI capabilities.
SIGNIFICANT · Mastodon — mastodon.social · 1d · [3 sources] · MASTO

Hermes Agent Hits 140K GitHub Stars, Nvidia RTX as Local Inference Bedrock Hermes Agent hit 140K GitHub stars, most-used on OpenRouter. Runs locally on Nvidia R

NHN has launched a substantial 7,656-GPU cluster in Seoul, South Korea, aimed at domestic enterprise AI workloads, positioning itself against competitors like Naver and Kakao. Meanwhile, the Hermes Agent project has surpassed 140,000 GitHub stars, demonstrating its popularity for local AI inference, particularly on Nvidia RTX hardware using models like Qwen 3.6. Separately, a new arXiv paper introduces DPAA, a framework designed to reduce bias in Graph Neural Network (GNN) based recommender systems by adaptively reweighting message passing. AI

IMPACT NHN's large GPU cluster boosts domestic AI capacity, while Hermes Agent's popularity highlights local inference trends and the DPAA paper offers a new method for debiasing recommender systems.
RESEARCH · Mastodon — mastodon.social · 1d · MASTO

Though Google Cloud Next in Las Vegas was a couple weeks ago, I'm still working through it and trying to process everything I learned there. Three days, 32,000

Google Cloud has introduced a new agent accountability infrastructure designed to provide cryptographic IDs and auditable action trails for AI agents. This system includes real-time anomaly detection to flag unusual agent reasoning and trace it back to its source. GE Appliances is already deploying 800 AI agents using this technology for manufacturing and supply chain operations, enabling autonomous decision-making without direct human oversight. AI

IMPACT Enhances enterprise AI governance by providing auditable trails and anomaly detection for autonomous agents.
SIGNIFICANT · TechCrunch AI Italiano(IT) · 1d · [6 sources] · MASTO

Wirestock raises $23M to supply creative multi-modal data to AI labs

Wirestock has secured $23 million in Series A funding to expand its business of supplying multi-modal data to AI labs. The company pivoted in 2023 to provide datasets of images, videos, and 3D content, working with over 700,000 artists and designers. This funding will be used to hire for research, engineering, and product roles, and to build enterprise software for AI labs. Wirestock currently supplies data to six major foundation model makers and has an annual revenue run-rate of $40 million. AI

IMPACT Accelerates the availability of diverse, multi-modal data crucial for training more capable foundation models.
RESEARCH · Mastodon — fosstodon.org Polski(PL) · 1d · MASTO

Poland ceases to be an importer of foreign military technologies. The Polish Army debuts BMS Legion, its own battlefield management system, which integrates

Poland has introduced its indigenous battlefield management system, BMS Legion, for its military. This system integrates K2GF tanks into a combat network, marking a strategic shift away from relying on foreign military technology. The distributed system incorporates the AIRON TACTIC AI and ensures full control over its source code. AI

IMPACT Enhances military operational capabilities through integrated AI for battlefield management.
RESEARCH · Mastodon — mastodon.social Čeština(CS) · 1d · MASTO

China is heading to the Moon with a humanoid robot and artificial intelligence. The Chang'e 8 mission will launch in 2029 to test advanced robotics in a hostile environment.

China is planning a lunar mission in 2029 named Chang'e 8, which will test advanced robotics and artificial intelligence in the harsh lunar environment. This mission is part of China's broader ambitions for future space exploration. AI

IMPACT Demonstrates AI's role in future extraterrestrial exploration and robotic capabilities.
RESEARCH · The Decoder · 1d · [3 sources] · MASTO

Alibaba's Qwen-Image-2.0 doubles compression and cuts generation steps from 40 to 4

Alibaba has released Qwen-Image-2.0, an image generation model that significantly improves efficiency. The model doubles image compression rates compared to competitors and reduces generation steps from 40 to just 4. It also incorporates a module for expanding user prompts into more detailed descriptions, and currently ranks 9th on the LMArena leaderboard. AI

IMPACT Enhances efficiency in image generation models, potentially lowering computational costs for AI applications.
SIGNIFICANT · Mastodon — mastodon.social Türkçe(TR) · 1d · [2 sources] · MASTO

📰 AI Image Model Training in 2026: Top Tools and Strategies AI image model training is accelerating with LoRA and video-based solutions. SeaArt

Alibaba has released its new Qwen-Image-2.0 model, significantly reducing generation steps from 40 to 4 and doubling image compression. This advancement also includes automatic enhancement of user prompts. The model has achieved a notable ranking of 9th on the LMArena blind comparison platform. AI

IMPACT Significantly reduces image generation time and improves compression, potentially accelerating creative workflows.
SIGNIFICANT · X — MiniMax AI · 1d · X

Great conversations with the dAI team & excited for what’s next!

MiniMax AI has announced a new model, though details are scarce. The company shared a brief statement on X, expressing excitement about future developments and interactions with their "dAI team." Specifics regarding the model's capabilities or release date were not provided in the announcement. AI

IMPACT New model releases from frontier labs often signal shifts in AI capabilities and competitive landscapes.
RESEARCH · Mastodon — fosstodon.org · 1d · MASTO

Electronics component linchpin TDK is prepared to add to what is already its biggest capital spending campaign ever in a push to ride the global boom in generat

TDK, a major electronics component manufacturer, is increasing its capital spending to meet the surging demand driven by the generative AI boom. The company is already undertaking its largest-ever investment campaign and is ready to allocate further funds. This expansion is aimed at capitalizing on the global growth in AI technologies. AI

IMPACT TDK's increased investment in component manufacturing will likely support the scaling of AI infrastructure and model development.
SIGNIFICANT · Mastodon — fosstodon.org 日本語(JA) · 1d · [3 sources] · MASTO

Google to Save 6 Million Teachers Nationwide with AI! "Google AI Educator Series" Launches. Restoring "Original Education" Where Teachers Face Students https://www.yayafa.com/2800387/ #AgenticAi #AI #ArtificialGeneralIntelligence

Anthropic has launched Claude for small businesses, offering support for sales strategies and billing. Separately, major Japanese financial institutions including Mitsubishi UFJ, Mizuho, and Sumitomo are gaining access to Claude Mythos, indicating its advanced capabilities. Meanwhile, Google is initiating an AI Educator Series aimed at assisting six million teachers across the US, allowing them to focus more on direct student interaction. AI

IMPACT New AI tools and services are being released for small businesses, financial institutions, and educators, indicating broader adoption across diverse sectors.
SIGNIFICANT · Tom's Hardware · 1d · [2 sources] · MASTO

Leading-edge foundry roadmaps for TSMC, Intel and Samsung — outlining the path to 1.4nm nodes and beyond

TSMC, Intel, and Samsung are all producing chips using 2nm-class process technology, with Samsung and TSMC having started mass production in mid-to-late 2025. Their future roadmaps diverge significantly: TSMC prioritizes predictable scaling and specialization, Samsung focuses on yield improvement with iterative node variants, and Intel pursues an aggressive strategy with advanced technologies like gate-all-around transistors and backside power delivery, aiming for 1.4nm nodes by 2027-2028. AI

IMPACT Advanced chip manufacturing progress directly impacts the performance and availability of AI hardware, influencing future AI development and deployment.
RESEARCH · Mastodon — fosstodon.org · 1d · MASTO

Cerebras set for debut in stock market gripped by #AI mania - #stocks www.reuters.com/legal/transa... Cerebras set for debut in stoc...

Cerebras Systems, a company specializing in AI hardware, is preparing for its initial public offering amidst a surge of investor interest in artificial intelligence. The company, known for its wafer-scale chips designed for AI computation, is expected to leverage the current market enthusiasm for AI-related stocks. This move signals a growing trend of AI infrastructure companies seeking public market validation. AI

IMPACT The IPO of Cerebras Systems could signal increased investment in AI hardware infrastructure, potentially accelerating the development and deployment of AI technologies.
RESEARCH · Mastodon — fosstodon.org 日本語(JA) · 1d · [3 sources] · MASTO

NVIDIA Q1 Fiscal Year 2027 Earnings Call – moomoo Community https://www.yayafa.com/2800377/ # AgenticAi # AI # ArtificialGeneralIntelligence # ArtificialIntelligence # NVIDIA # エージ

A new multimodal AI model named "Llama 4 Behemoth" has been announced, with its name suggesting immense capabilities. Separately, Nvidia held its Q1 fiscal year 2027 earnings call. The specific details and performance metrics of Llama 4 Behemoth are not yet fully disclosed, but its designation implies a significant advancement in AI technology. AI

IMPACT The announcement of a new multimodal AI model suggests potential advancements in AI capabilities, while Nvidia's earnings call provides insights into the financial health of a key player in the AI hardware market.
RESEARCH · Mastodon — fosstodon.org Polski(PL) · 1d · [2 sources] · MASTO

Breakthrough! Google stops AI-created 'zero-day' exploit for the first time. This is a milestone in cybersecurity! What does this mean for your protection?

Google has successfully blocked a zero-day exploit that was reportedly developed using artificial intelligence. This marks a significant advancement in cybersecurity, demonstrating AI's dual role in both creating and defending against sophisticated cyber threats. The development highlights the evolving landscape of digital security as AI capabilities continue to advance. AI

IMPACT Demonstrates AI's growing role in both creating and defending against sophisticated cyber threats, marking a key moment in cybersecurity.
RESEARCH · Mastodon — fosstodon.org Polski(PL) · 1d · MASTO

Luma's generative model Uni-1.1 is now available as an API, moving beyond closed ecosystems. With high performance in benchmarks and aggressive

Luma AI has released its Uni-1.1 generative model via an API, aiming to provide an alternative to closed-ecosystem solutions from major tech companies. The model demonstrates strong performance in benchmarks and features competitive pricing. This move positions Uni-1.1 as a potential challenger in the generative AI landscape. AI

IMPACT Provides a new API-based generative model option, potentially increasing competition and offering alternatives to closed ecosystems.
SIGNIFICANT · Mastodon — sigmoid.social · 1d · [4 sources] · MASTO

Market Currents Surge on AI Enthusiasm, Trade Clouds Linger US stock futures rise on AI excitement, S&P 500 and Nasdaq reach new records. Investors watch US-Chi

The US stock market, including the S&P 500 and Nasdaq, has reached record highs, fueled by significant investments in AI by major tech companies. This surge in market performance is occurring despite broader economic challenges such as rising inflation and global conflicts. Investor attention is also focused on upcoming Federal Reserve speeches and US-China trade talks. AI

IMPACT AI investments are driving record highs in major US stock indices, signaling strong market confidence in the sector's growth potential.
SIGNIFICANT · 36氪 (36Kr) 中文(ZH) · 1d · [26 sources] · MASTO

Polls show over 70% of Americans oppose building AI data centers near their homes

A recent Gallup poll indicates that 71% of Americans oppose the construction of AI data centers near their homes, a sentiment stronger than their opposition to nuclear power plants. Key concerns include the significant water and energy consumption, pollution, and potential increases in utility costs associated with these facilities. This widespread public resistance is beginning to impact data center development, with some projects facing delays, scaling back, or outright cancellation. AI

IMPACT Widespread public opposition to AI data centers could significantly slow down the expansion of necessary compute infrastructure.
RESEARCH · Mastodon — fosstodon.org · 1d · [2 sources] · MASTO

How humanoids learn to read the room - Via @therobotreport #AI #ArtificialIntelligence 💻 🧠 #Robotics 🤖 How humanoids learn to read th...

Humanoid robots are being developed with advanced AI capabilities to better understand and respond to social cues. Researchers are focusing on enabling these robots to interpret human emotions and intentions, allowing for more natural and intuitive interactions. This development aims to improve the integration of robots into human environments and enhance their ability to perform tasks requiring social awareness. AI

IMPACT Enhances the potential for robots to interact naturally with humans, improving their utility in social and service roles.
RESEARCH · Mastodon — fosstodon.org · 1d · [6 sources] · MASTO

🤖 Anthropic's Claude Mythos AI: Cybersecurity Savior or Pandora's Box? Anthropic's Claude Mythos AI, still under wraps, claims it can detect browser vulnerabili

Anthropic's new AI model, Claude Mythos, is demonstrating advanced capabilities in cybersecurity, including the ability to detect thousands of browser vulnerabilities. This rapid evolution has raised concerns among cybersecurity experts and prompted the formation of a public-private task force in Japan to address AI-driven cyber threats. The model's potential impact has spurred urgent policy discussions and defensive initiatives. AI

IMPACT New AI models with advanced cybersecurity detection capabilities could accelerate the development of both offensive and defensive cyber tools.
SIGNIFICANT · TechCrunch AI · 1d · [7 sources] · MASTO

Who decides what AI tells you? Campbell Brown, once Meta’s news chief, has thoughts

Campbell Brown, formerly Meta's news chief, has launched Forum AI to assess how large language models handle complex, high-stakes topics like geopolitics and mental health. The company enlists prominent experts, including former Secretary of State Tony Blinken and historian Niall Ferguson, to develop benchmarks and train AI judges. Brown expressed concern that AI companies prioritize coding and math over accuracy, leading to biased or incomplete information, and aims to push for greater truthfulness in AI outputs, particularly for enterprise applications where liability is a concern. AI

IMPACT Establishes a new standard for AI evaluation, potentially influencing model development and enterprise adoption by focusing on accuracy and truthfulness.
RESEARCH · MarkTechPost · 2d · [4 sources] · MASTO

Nous Research Releases Token Superposition Training to Speed Up LLM Pre-Training by Up to 2.5x Across 270M to 10B Parameter Models

Nous Research has developed Token Superposition Training (TST), a new method designed to significantly accelerate the pre-training of large language models. This technique can reduce pre-training time by up to 2.5 times for models ranging from 270 million to 10 billion parameters, without altering the model's architecture or how it performs inference. TST achieves this by modifying the training loop in two phases: an initial 'superposition' phase where token embeddings are averaged and processed in larger bags, followed by a 'recovery' phase that reverts to standard training. Experiments showed TST achieving lower final training loss with substantially less compute time compared to traditional methods. AI

IMPACT Accelerates LLM pre-training, potentially reducing compute costs and time for developing new large language models.
SIGNIFICANT · Mastodon — sigmoid.social 한국어(KO) · 2d · [6 sources] · MASTO

HMRC to use AI from British tech firm Quantexa to spot fraud and tax return errors, signing a 10-year, £175 million contract to implement AI technology for detecting tax fraud and reporting errors.

The UK's tax authority, HMRC, has awarded a 10-year, £175 million contract to the British AI firm Quantexa. This partnership will leverage Quantexa's AI technology to enhance the detection of tax fraud and errors in tax returns. The system is designed to combine internal and external data to identify fraudulent networks, with a built-in human review process for automated decisions to ensure transparency. AI

IMPACT This contract signifies a significant government investment in AI for public services, potentially setting a precedent for other agencies in fraud detection and data analysis.
RESEARCH · arXiv cs.CL · 2d · [5 sources] · MASTO

Herculean: An Agentic Benchmark for Financial Intelligence

A new benchmark called Herculean has been developed to assess the financial intelligence of AI agents, revealing that current frontier models struggle with complex tasks like hedging and auditing. This highlights a significant gap in their ability to translate reasoning into dependable workflow execution in high-stakes financial scenarios. Concurrently, the financial services industry is emphasizing the critical need for robust data readiness to support agentic AI, as regulatory demands and the complexity of financial data require accessible, reliable, and governed data stores. AI

IMPACT Highlights a critical gap in current AI agent capabilities for complex financial workflows, emphasizing the need for better data governance and model execution.
SIGNIFICANT · 36氪 (36Kr) 中文(ZH) · 2d · [3 sources] · MASTO

TSMC expects the global chip market to exceed $1.5 trillion by 2030

Taiwan Semiconductor Manufacturing Company (TSMC) projects the global semiconductor market will surpass $1.5 trillion by 2030, a significant increase from previous forecasts. The company anticipates that artificial intelligence and high-performance computing will be the primary drivers, accounting for 55% of this market. To meet this demand, TSMC is accelerating its capacity expansion, with plans for new fabrication plants and advanced packaging facilities in 2026. AI

IMPACT This forecast highlights AI's critical role in driving demand for advanced semiconductor manufacturing, signaling continued investment and expansion in the sector.
SIGNIFICANT · TechCrunch AI · 2d · [3 sources] · MASTO

Cisco cuts nearly 4,000 jobs to spend more on AI, reports ‘record quarterly revenue’

Cisco Systems is reducing its workforce by nearly 4,000 employees, approximately 5% of its staff, despite achieving record quarterly revenue of $15.84 billion. The networking giant is reallocating resources to invest heavily in AI infrastructure hardware and cybersecurity. This strategic shift aims to capitalize on exploding hyperscaler demand and address ongoing security vulnerabilities. AI

IMPACT Cisco's pivot to AI infrastructure hardware signals a major shift in enterprise networking, potentially impacting the supply chain for AI development.
COMMENTARY · Mastodon — sigmoid.social · 30m · MASTO

🦊😈🐾🕵️💸 5 takeaways about NPR's reporting on the whistleblower report about DOGE at the NLRB # AI Q: 🕵️ Risk vs efficiency? 🛡️ Data Privacy | ✊ Labor Rights | ⚖️

NPR's reporting on a whistleblower's allegations regarding DOGE at the NLRB raises questions about risk versus efficiency. The report touches upon critical areas such as data privacy, labor rights, and government oversight. These elements highlight the complex interplay between technological advancements and regulatory frameworks. AI

IMPACT Discusses the intersection of technology, labor rights, and government oversight, relevant for understanding AI's broader societal impact.
COMMENTARY · Mastodon — sigmoid.social · 1h · [2 sources] · MASTO

Post 2 — the response: Violence is what happens when people have no other option that feels real. A constitutional process that feels accessible and genuine red

Technology can enable genuine national deliberation by overcoming logistical barriers, potentially making constitutional processes more accessible and reducing societal pressure. Federated platforms and resilient mesh infrastructure are key to building such systems. This approach could offer a viable response to AI-driven economic displacement by fostering equitable benefit distribution and accessible governance. AI

IMPACT The cluster discusses the potential societal impact of AI on economic displacement and governance, suggesting technological solutions for equitable benefit distribution and accessible deliberation.