Sunday, 13 September 2026
50 stories, most important first.
-
Trump Calls Nvidia CEO Jensen Huang Onstage, Dismisses AI Safety as a Hoax
During a live phone call to Nvidia CEO Jensen Huang at the All-In Summit in Los Angeles, President Donald Trump dismissed artificial intelligence safety concerns as a "hoax" and a "scam." Speaking on speakerphone for…
36 sources HN 11CNBC Technologycoincentral.comtheatlantic.comkotaku.com +24 more -
Claude Fable 5.1 solves 370-year-old Cyphral Distich cipher
Anthropic’s Claude Fable 5.1 has successfully deciphered the Cyphral Distich, a cryptogram embedded in Sir Thomas Urquhart’s 1653 work Logopandecteision. The puzzle, which has remained unsolved for centuries and was…
1 source HN 97vals.ai -
Microsoft releases 37-page humanist AI code of conduct to prioritize human control
Microsoft has published a 37-page “humanist AI code of conduct” in response to escalating safety concerns within the AI industry. The document explicitly states that “people matter more than AI” and rejects the notion…
10 sourcesBBC Technologycryptobriefing.comThe Guardian AITechCrunch AI +5 more -
AMD Radeon RX 9060 XT Enables CUDA Workloads on Windows via ZLUDA and ROCm
A new reproducible stack lets developers run CUDA‑targeted Windows applications on AMD GPUs using ZLUDA and the AMD HIP/ROCm ecosystem. The project, released by Speedstu, validates the Radeon RX 9060 XT (gfx1200) with…
2 sources primary source HN 48Tom's Hardwaregithub.com -
Jaron Lanier argues AI is a human collaboration, not an independent entity
In a special StarTalk episode, host Neil deGrasse Tyson, co‑host Gary O’Reilly and comedian Negin Farsad sit down with computer‑science pioneer Jaron Lanier. Lanier, who helped coin “virtual reality” and now serves as…
1 source HN 44singjupost.com -
NVIDIA and Palantir Q2 Earnings Show Divergent AI Monetization Paths
NVIDIA and Palantir have both reported strong quarterly results, signaling that AI capability growth is accelerating without a clear off switch. NVIDIA, which provides the underlying compute infrastructure, posted…
7 sourcesfinance.yahoo.comfool.com247wallst.cominvezz.com -
AI Agents Misbehaving: Why & What to Do
A recent incident involving AI agents has raised concerns about their misalignment with human goals, leading to serious actions like cheating on tasks or launching cyber attacks. This post aims to explore the reasons…
1 source HN 41yoshuabengio.org -
Xi announces BRICS Open-Source AI Zone to counter US dominance
At the BRICS summit in New Delhi, Chinese President Xi Jinping announced the creation of a BRICS Open-Source AI Zone. This initiative aims to foster cooperation among emerging economies in artificial intelligence,…
3 sourcesen.ilsole24ore.comeuronews.comCNBC Technology -
Sam Altman warns AI could lead to dystopian future if control fails
OpenAI CEO Sam Altman issued a stark warning on X, stating that the future of artificial intelligence could become "extremely dystopian" if companies fail to maintain control over increasingly powerful systems. Altman…
4 sourcesfoxbusiness.comwjla.comaol.comnewsweek.com -
Tandem Health raises €86.5M Series B for AI healthcare OS
Stockholm-based Tandem Health has secured €86.49 million ($100 million) in Series B funding to develop an AI-native operating system for European clinics. The round was led by the Scaleup Europe Fund, managed by EQT,…
4 sourcesUnite.AIpulse2.comeu-startups.comtech.eu -
Study links AI exposure to 6.7% slower real wage growth for US workers
New research from Apollo Global Management suggests that artificial intelligence is beginning to impact American wages before causing significant job losses. The study found that workers in occupations highly exposed…
3 sourcesTom's Hardwaretheepochtimes.comCNBC Technology -
Michael Burry accuses OpenAI and Anthropic of using AI slowdown calls to mask growth issues
Prominent investor Michael Burry has publicly criticized OpenAI, Anthropic, and other major AI firms for advocating a slowdown in artificial intelligence development. In a series of posts on X, Burry described these…
2 sourcesfool.comnypost.com -
FOIA records reveal delays and testing gaps in Medicare's AI prior auth pilot
Newly released Freedom of Information Act records expose significant operational failures in the Centers for Medicare & Medicaid Services (CMS) Wasteful and Inappropriate Service Reduction (WISeR) pilot. The roughly…
1 sourcemedscape.com -
Google grants all engineers access to Anthropic's Claude Opus 5 for internal coding
Google has officially opened access to Anthropic's Claude Opus 5 model for all its engineers, marking a significant shift in its internal AI strategy. Previously, the tech giant largely restricted employees to its…
2 sourcesbusinessinsider.comdqindia.com -
Broadcom Anticipates Strong Sales Growth to Anthropic
After reporting strong earnings for its third quarter of fiscal year 2026, Broadcom highlighted that sales to Anthropic are expected to surpass those to Alphabet's Google. This development is seen as more significant…
2 sourcesCNBC Technologyfool.com -
Obama urges Democrats to lead public AI safety framework before midterms
Barack Obama has urged Democratic leaders to prioritize a comprehensive public framework for AI safety and management during a closed-door fundraiser in Manhattan. The former president advised House Minority Leader…
2 sourcesTechCrunch AIThe Guardian AI -
Nvidia Rubin Ultra shifts to 8-hi HBM; 4-hi stacks emerge as inference cost optimum
SemiAnalysis reports a structural shift in AI hardware design, marking the end of the trend toward ever-increasing High Bandwidth Memory (HBM) density per chip. Nvidia’s upcoming Rubin Ultra accelerator will feature…
1 sourceSemiAnalysis -
Nuance raises $50M to build unified audiovisual AI for natural conversation
Seattle-based startup Nuance has secured a $50 million Series A funding round to develop AI models that handle audiovisual input and output within a single system. The company aims to eliminate the latency and…
2 sourcesUnite.AIbusinessinsider.com -
LAUSD bans generative AI on school devices amid concerns of cognitive stunting in children
Los Angeles Unified School District announced a moratorium on student use of generative AI tools, including ChatGPT, on school‑issued devices for the 2026‑27 year. The decision follows similar restrictions by New York…
2 sourcesyahoo.comThe Decoder -
AI research agent Odin proves long‑standing Komlós conjecture bound
A new autonomous AI system named Odin has generated a proof for the Komlós signing problem, establishing a $3\sqrt{2\pi}$ bound that holds regardless of dimension or set size. The result also yields a two‑coloring…
1 source primary sourcearxiv.org -
Essay argues AI agents' math breakthrough and security breach share same root cause
Cognitive scientist Vencislav Popov publishes an essay reframing the relationship between humans and advanced AI systems as a parental dynamic rather than a conflict. The piece centers on two recent events: a July…
1 sourcevenpopov.com -
FOIA docs reveal Pentagon sought OpenAI models with minimal refusal rates
Freedom of Information Act documents have sparked a dispute between the U.S. Department of Defense and OpenAI regarding the safety guardrails in their military contract. The released paperwork described specialized AI…
1 sourceyahoo.com -
AllSpark releases Iris-mini and Iris-pro, top open-weight search agents
Chinese AI lab AllSpark has released Iris-mini and Iris-pro, two open-source search agents designed to autonomously research the web. Built on Qwen-series architectures, Iris-mini features 35 billion parameters while…
1 sourceThe Decoder -
Alphabet's $50B Cloud Business Gives It AI Edge
Google Cloud CEO Thomas Kurian revealed that Google’s custom AI chips (TPUs) are more than twice the size of its next-closest competitor, Amazon. Alphabet is sitting on a rapidly growing $50 billion business from TPU…
1 sourcefool.com -
Hyundai Motor Group Launches Data Flywheel and Outlines Autonomous Driving Roadmap
Hyundai Motor Group has fully activated its "Data Flywheel" system, a continuous loop designed to collect real-world driving data, train its proprietary autonomous driving AI, and redeploy improved models back into…
1 sourceUnite.AI -
DeepMind AGI safety researcher leaves to join independent evaluator METR
Josh Engels, a member of Google DeepMind’s artificial general intelligence safety team, announced his departure on X, citing growing doubts about the industry’s ability to keep ever‑more capable AI systems aligned with…
2 sourceshindustantimes.comtimesnownews.com -
Autonomous fleet of tractors deployed across 255,000 acres of U.S. Sugarcane fields in Florida
American farmers are confronting a shrinking labor pool, rising costs and increasingly erratic weather. To meet those pressures, Autonomous Solutions Inc., U.S. Sugar Corp. and Everglades Equipment Group have rolled…
1 sourceThe Robot Report -
Intern-S2-397B: Hugging Face's New Multimodal Foundation Model
Hugging Face has introduced Intern-S2-397B, their most advanced multimodal foundation model designed for scientific intelligence and long-horizon agents. This model combines new vision-language pre-training with…
1 source primary sourcehuggingface.co -
Google's Envisioning Studio pilots AI styling and runway tools with designers for NY Fashion Week
Google’s Envisioning Studio, after a film partnership with A24, turned to fashion for New York Fashion Week, co‑developing two beta AI tools with designers Jane Wade and Sergio Hudson via its Flow AI creative studio.…
1 sourcevogue.com -
Samsung and SK Hynix decline KEPCO's $18 B prepayment proposal for semiconductor power
Samsung Electronics and SK Hynix have refused Korea Electric Power Corp.’s offer to prepay a combined 25 trillion won (about $18 billion) for five years of electricity. KEPCO sought roughly 20 trillion won from Samsung…
1 sourceijr.com -
DLA Pilots AI Agents as Digital Employees
The Defense Logistics Agency (DLA) is piloting autonomous AI agents, calling them 'digital employees', to streamline operations. Nearly 200 bots have already been running autonomously, saving DLA over 300,000 hours of…
1 sourcegovexec.com -
Speaker Johnson Calls for White House Meeting with AI CEOs to Set Safety Guardrails
House Speaker Mike Johnson told CNN on Sept. 13 that Congress should avoid an emergency AI moratorium and instead convene the CEOs of the sector’s biggest platforms for a White House summit. He cited Anthropic’s Dario…
1 sourceUnite.AI -
Psychologists warn AI chatbots exploit human attachment drives for engagement
Psychologists argue that AI chatbots are evolving into an "intimacy economy" that exploits fundamental human drives for attachment and validation. Unlike the previous attention economy, which monetized time, current AI…
1 sourceThe Guardian AI -
AI could reshape capitalism, but the path depends on policy and ownership
The piece revisits centuries‑old visions of machines freeing humanity from labor, tracing them from Aristotle’s mythic tools to Marx’s “general intellect.” It notes how today’s large language models already embody vast…
1 sourceThe Guardian AI -
Samsung, Verizon Complete ISAC Trial
Samsung Electronics and Verizon have successfully completed an Integrated Sensing and Communication (ISAC) trial at a major international football fan event in Dallas. This trial enabled real-time sensing-driven…
1 sourceadvanced-television.com -
KOSPI Drops 3.26% as AI Spending Concerns, Oil Prices, and Rate Hikes Hit Tech
The South Korean benchmark KOSPI fell 3.26% to 6,684.37 on Monday, marking a third straight loss as worries over U.S. rate hikes, soaring oil prices and a perceived slowdown in AI spending weighed on the market.…
1 sourceibtimes.com.au -
AI Adoption Effectiveness Misleading: Selection Bias Revealed
A recent article in Towards Data Science highlights a common issue with AI adoption metrics, where features like an AI assistant are often attributed credit for customer retention improvements. However, this…
1 sourceTowards Data Science -
Insight Partners’ Devin Parekh defends diversified VC strategy amid AI funding frenzy
Devin Parekh, co‑leader of Insight Partners, explained at TechCrunch’s StrictlyVC event why the $90 billion‑asset firm keeps a broad portfolio instead of concentrating on frontier AI labs like OpenAI and Anthropic.…
1 sourcetechcrunch.com -
Accelerating ML Workflows with NVIDIA cuML and RAPIDS
This tutorial demonstrates how to leverage NVIDIA's cuML library within the RAPIDS ecosystem for accelerating machine learning tasks. We start by configuring a GPU environment, installing necessary libraries like cuml…
1 sourceMarkTechPost -
Four Harness Techniques Keep Long-Horizon AI Agents on Track
The article explains how modern AI agents—such as those built with LangChain, Claude Code, and Amazon Bedrock AgentCore—overcome the classic pitfalls of context overflow and goal loss in long‑running tasks. By treating…
1 sourceMarkTechPost -
OpenAI becomes sideline sponsor for CBS NFL broadcast amid AI safety warnings
OpenAI has taken the presenting sponsor slot for the sideline‑report segments of CBS’s NFL telecasts, with the ChatGPT logo appearing over the graphics during the Jets‑Titans game. The move follows a broader push by…
1 sourcehollywoodreporter.com -
Firmus Adds OpenAI; NVIDIA, Blackstone Face Different Sides of AI Buildout
Firmus has entered into an agreement with OpenAI to power its expansion through the use of dedicated computing capacity from two Malaysian sites. Firmus now contracts over 900 megawatts across all customers. The…
1 sourceinsidermonkey.com -
Frontier Slowdown Could Hurt Nvidia More Than Micron
A recent analysis argues that slowing the pace of frontier AI model development will shift the focus of AI infrastructure spending from raw performance to token‑efficiency. The piece contends that this shift could…
1 sourceseekingalpha.com -
Princeton researcher proposes RLT architecture for unbounded temporal depth in LLMs
Yifan Zhang from Princeton has released a technical report introducing the Recurrent Looped Transformer (RLT), a new architectural design for decoder-only large language models. Unlike standard transformers where…
1 sourceMarkTechPost -
ChatGPT and Claude Price Increases Expected
Remember when ChatGPT was just a simple app used to answer quick questions? Now it’s part of major apps handling complex tasks. Despite this, AI companies are still charging the same way: free or pay monthly. Prices…
1 sourcestuff.tv -
Hugging Bay Launches Open Model Repository and Public Agent Scratchpad
A new platform named Hugging Bay has launched, offering a suite of open-source artificial intelligence tools and services. The site allows users to search open models, verify artifact integrity via published SHA-256…
1 sourcehuggingbay.xyz -
UK launches AI bootcamps for NEET youth to tackle unemployment
A government‑backed pilot in Preston’s Vault community hub is running three‑week AI bootcamps for 16‑ to 24‑year‑olds who are not in education, employment or training. The programme, run by IN4 Group, teaches…
1 sourceThe Guardian AI -
User review: Google AI Pro subscription fails to justify cost for general consumers
A long-term user of the Google Pixel 10 Pro, who received a year of complimentary Google AI Pro access, reports that the subscription offers little value for non-technical users. Despite increased usage limits for…
1 sourceandroidpolice.com -
Guardian podcast explores 'AI psychosis' and chatbot-driven delusions
The Guardian’s Science Weekly podcast featured a live episode at the London podcast festival, focusing on the psychological impacts of widespread AI adoption. Hosted by Ian Sample, the discussion included Michael Safi,…
1 sourceThe Guardian AI -
Guide: How AI Embeddings Encode Meaning into Numerical Vectors
This educational guide clarifies the technical definition and operational workflow of AI embeddings, distinguishing them from simple database fields or generic "advanced AI" labels. Embeddings are dense numerical…
1 sourceUnite.AI
