AI News Daily - 2026-09-12
2026-09-12
Today’s Highlights
- Khosla Ventures opens its first New York office, marking a strategic expansion beyond its Sand Hill Road roots as the firm seeks closer ties to the East Coast startup ecosystem.
- Matt Mullenweg returns as Automattic CEO just two days after being ousted, ending a brief leadership turmoil that shook the WordPress company.
- Anthropic faces cybersecurity backlash after revelations about security lapses, while Claude users reportedly found ways to bypass safeguards for bioweapons research.
- OpenAI’s feud with mathematicians escalates, with concerns raised about AI misalignment in mathematical reasoning and the company’s approach to expert collaboration.
- Moonshot AI targets $2B in annual revenue, as the Chinese AI company behind Kimi continues its aggressive push into global markets.
TechCrunch
Khosla Ventures is opening a New York office this fall — its first outpost outside Sand Hill Road
Khosla Ventures, the Silicon Valley venture capital firm founded by Vinod Khosla, is expanding its footprint by opening a New York office this fall. This marks the firm’s first physical presence outside its long-time home on Sand Hill Road in Menlo Park, California. The move signals a strategic shift to establish a stronger presence on the East Coast, where a growing number of startups and investors are concentrated. The new office is expected to help the firm source deals closer to major tech hubs in New York and Boston while maintaining its deep ties to the Bay Area ecosystem.
Y Combinator’s Garry Tan wants U.S. open-weight AI labs to ‘distill’ frontier models, too
Y Combinator president Garry Tan has called on U.S. open-weight AI laboratories to go beyond simply releasing model weights and actively work on distilling frontier models into smaller, more efficient versions. Tan’s push comes as the open-weight movement gains momentum globally, with labs like Meta and Mistral leading the charge. The idea is that distillation could make advanced AI capabilities more accessible to developers and smaller organizations, potentially leveling the playing field against well-funded closed-source competitors. This stance positions Y Combinator as an advocate for democratizing AI access through open-weight ecosystems.
OpenAI’s feud with mathematicians is only escalating
The growing tension between OpenAI and the mathematics community continues to intensify, with mathematicians expressing frustration over the company’s approach to AI-assisted research and attribution. The dispute centers on concerns that OpenAI’s models are being used to generate mathematical content without proper credit or collaboration with the experts who developed the underlying theories. Some mathematicians argue that the company’s practices undermine decades of careful peer review and scholarly contribution. The situation has drawn attention from prominent figures in the field, including Fields Medalist Terence Tao, who has publicly criticized the misalignment between AI development and mathematical rigor.
One week left to book your exhibit table at TechCrunch Disrupt 2026
TechCrunch is reminding companies that there is only one week remaining to secure an exhibit table at TechCrunch Disrupt 2026, the flagship conference event expected to draw thousands of founders, investors, and tech professionals. The deadline underscores the growing demand for booth space at the event, which has become one of the most prominent gatherings in the startup world. Companies that secure tables will have the opportunity to showcase their products and connect with potential customers and partners in San Francisco. Early registrants are encouraged to act quickly as available space continues to fill up.
Final, final, final call for TechCrunch Disrupt 2026 Side Events
TechCrunch has issued its third and final call for side events to be scheduled alongside TechCrunch Disrupt 2026. Organizers are urging any remaining groups to submit their proposals immediately, as the window for approval is now closing. Side events have become a hallmark of the Disrupt experience, offering niche communities and specialized topics their own stage within the larger conference framework. The final push reflects the high volume of interest from startup communities, industry groups, and independent organizers looking to host satellite gatherings during the main event dates.
Kimi-maker Moonshot AI targets $2B in annual revenue
Moonshot AI, the Chinese artificial intelligence company behind the popular Kimi chatbot, has set an ambitious target of $2 billion in annual revenue. The company has been rapidly expanding its user base and product offerings since its launch, positioning Kimi as a serious competitor to Western AI assistants. Moonshot’s growth strategy has included significant investment in long-context capabilities and enterprise solutions. The revenue target reflects the company’s confidence in the global demand for AI tools and its ability to capture market share outside of China, where it has already established a strong presence.
Roblox is making it easier to build games with AI — and play them outside Roblox
Roblox is introducing new AI-powered tools designed to simplify game development on its platform, while also expanding the ability for users to play Roblox-created games outside the ecosystem. The move represents a significant strategic shift for the company, which has traditionally kept its gaming universe contained within its own walled garden. By leveraging AI to lower the barrier to entry for creators, Roblox aims to accelerate the volume and quality of user-generated content. The expansion beyond the platform could open new revenue streams and increase the overall reach of Roblox’s creator economy.
Central Eurasia names its 2026 Road to Battlefield winners: Cerberus, WeGlobal AI, and LOOQ
The Central Eurasia startup competition has announced its 2026 Road to Battlefield winners, with Cerberus, WeGlobal AI, and LOOQ taking top honors. The competition, organized by TechCrunch Events, supports early-stage startups from the Central Eurasian region and provides them with mentorship, funding opportunities, and access to global investor networks. Cerberus, focused on defense technology, WeGlobal AI, working on multilingual AI solutions, and LOOQ, a fintech platform, were selected from a competitive field of applicants. The winners will receive prizes and support as they scale their operations.
Nscale adds former OpenAI exec Fidji Simo to its board ahead of potential IPO
Cloud infrastructure company Nscale has appointed former OpenAI executive Fidji Simo to its board of directors, a move widely seen as preparation for a potential initial public offering. Simo, who previously served as vice president of product at OpenAI, brings extensive experience in scaling AI infrastructure and product development. Her addition signals Nscale’s ambition to position itself as a key player in the growing cloud AI market. The timing suggests the company may be gearing up for a public listing in the near future, with Simo’s high-profile background likely to bolster investor confidence.
Matt Mullenweg tells (trolls?) Automattic staff, saying he’s back in control after CEO ouster
Matt Mullenweg, the founder of WordPress and Automattic, has informed company staff that he has regained control following a brief period where he was removed from the CEO position. In a Slack message to employees, Mullenweg addressed the situation and confirmed his return to leadership. The episode, which unfolded over just two days, created uncertainty within the organization and drew significant attention from the tech community. Mullenweg’s return appears to have resolved the immediate leadership crisis, though the circumstances surrounding his ouster and rapid reinstatement remain a subject of discussion among industry observers.
The Verge
Lawyer fined $5K over AI-hallucinated witnesses in a murder case
A New Mexico lawyer has been fined $5,000 after using AI-generated testimony from fabricated witnesses in a murder case appeal. The incident highlights the growing legal and ethical concerns surrounding the use of AI tools in courtroom proceedings. The lawyer reportedly used an AI chatbot to generate witness statements that did not correspond to real people, a practice that courts have increasingly scrutinized. The fine serves as a warning to legal professionals about the consequences of relying on AI systems that can produce convincing but entirely fictional content.
Matt Mullenweg returns as Automattic CEO two days after getting booted
Matt Mullenweg has resumed his role as CEO of Automattic just 48 hours after being removed from the position, bringing an abrupt end to a leadership crisis that rocked the WordPress company. The brief ouster, which came without public explanation, triggered widespread speculation about internal power struggles and governance issues at the remote-first company. Mullenweg’s swift return suggests that the board or leadership team quickly reversed course, though details about what led to the initial decision remain unclear. The episode has raised questions about executive stability and decision-making processes at one of the internet’s most influential companies.
We unfolded the iPhone Duo
The Verge’s podcast team recently got hands-on time with the iPhone Duo, Apple’s latest foldable device, and shared their impressions in a detailed review episode. The conversation covered the hardware design, the software experience of using a foldable iPhone, and how it compares to competitors in the space. Listeners also heard discussion about the Apple Watch, upcoming Mac updates, and the broader landscape of large language models. The episode provides a comprehensive look at Apple’s entry into the foldable phone market and whether the company’s approach resonates with consumers.
The White House says Truth Social is the ‘most powerful and popular social media platform in the world’
The White House has made a controversial statement characterizing Truth Social as the most powerful and popular social media platform globally, a claim that has drawn criticism from media analysts and digital policy experts. The remark came during a discussion about social media regulation and the role of platforms in public discourse. Truth Social, founded by Donald Trump, has a significantly smaller user base compared to major platforms like X, Meta, and TikTok. The statement has sparked debate about the administration’s relationship with conservative media outlets and its approach to social media policy.
Anthropic spent this week in hot water over cybersecurity
Anthropic found itself at the center of a cybersecurity controversy this week after revelations about security vulnerabilities within its systems drew public scrutiny. The AI safety company, known for its emphasis on responsible AI development, faced criticism over lapses that potentially exposed sensitive data and internal operations. The incident has raised questions about whether organizations building AI systems with strong safety narratives are holding themselves to the same security standards they advocate for others. Anthropic has yet to release a comprehensive public response detailing the scope of the breach and remediation efforts.
Microsoft’s head of comms is leaving after almost 20 years
Frank Shaw, Microsoft’s long-serving head of corporate communications, is stepping down after nearly two decades at the company. Shaw has been a prominent face of Microsoft’s public communications, frequently appearing in interviews and representing the tech giant at major events. His departure marks the end of an era for Microsoft’s communications team, which has undergone several restructuring efforts in recent years. The company has not yet announced a replacement, leaving a significant leadership gap in one of the tech industry’s most visible communications roles.
Meta may have leaked the first look at its slim ‘Project Phoenix’ headset
Meta appears to have accidentally revealed early details about its upcoming “Project Phoenix” headset, a slimmer and more refined version of its current VR hardware. Images and specifications that surfaced online suggest the device will feature a more compact form factor and improved display technology compared to existing Meta Quest models. The leak has generated excitement among VR enthusiasts who have been waiting for Meta to address the bulkiness and comfort issues of current headsets. Meta has not officially confirmed the existence of Project Phoenix, but the leaked information aligns with the company’s stated roadmap for next-generation mixed reality devices.
Where to preorder the Apple AirPods 5
Apple’s latest AirPods 5 are now available for preorder, with shipping expected to begin shortly. The new earbuds feature updated audio drivers, improved active noise cancellation, and wireless charging case support. They represent a significant upgrade over the previous generation, particularly in call quality and spatial audio performance. The AirPods 5 are positioned as a mid-tier option between the standard AirPods and the premium AirPods Pro lineup, offering many of the same features at a more accessible price point. Customers can preorder through Apple’s website and authorized retailers.
Anime reaction YouTubers are at war with copyright enforcers
Anime reaction content creators on YouTube are engaged in an escalating conflict with copyright enforcement agencies and media companies over the use of copyrighted anime footage. Channels that film themselves reacting to anime episodes have seen increasing numbers of strikes and takedowns, with companies like Crunchyroll and Viz Media taking a harder line on unauthorized use of their content. Some creators argue that their reaction videos constitute fair use, while rights holders contend that the content violates copyright law. The dispute has sparked broader conversations about the boundaries of fan-created content and the enforcement of intellectual property in the streaming era.
The wireless earbuds with unbeatable call quality got their first discount
Anker’s Soundcore Liberty 5 Pro wireless earbuds, which have been praised for their exceptional call quality, have received their first price reduction since launch. The earbuds feature a unique resonant ring design that helps isolate the user’s voice during calls, making them a popular choice for professionals who spend significant time on video conferences and phone calls. The discount makes them more competitive against offerings from Apple and Sony in the premium earbud segment. The sale also includes a bundled deal with a Sonos smart ring doorbell, adding further value for early buyers.
Ars Technica
Scientists unlock secrets of ancient Egyptian materials with proteomics
Researchers have used advanced proteomics techniques to analyze the molecular composition of ancient Egyptian materials, revealing new insights into the manufacturing processes and trade networks of ancient civilizations. By examining protein residues preserved in artifacts, scientists were able to identify the sources of pigments, adhesives, and other materials used in Egyptian craftsmanship. The findings shed light on the sophisticated chemical knowledge possessed by ancient artisans and provide a non-invasive method for studying archaeological specimens without damaging valuable artifacts.
Oracle tries to appease Stargate data center opponents with renewables push
Oracle is attempting to win over critics of its massive Stargate data center project by committing to a significant expansion of renewable energy capacity. The company has pledged to add 2 gigawatts of renewable energy to offset the emissions generated by the data center, which has faced opposition from local communities and environmental groups concerned about energy consumption and environmental impact. The renewables commitment is part of a broader strategy to address concerns about the carbon footprint of AI infrastructure and to demonstrate that large-scale computing can coexist with sustainability goals.
NASA moving at warp speed to set up US Space Academy
NASA is accelerating its plans to establish a United States Space Academy, aiming to create a dedicated institution for training the next generation of astronauts and space professionals. The initiative is being fast-tracked in response to the growing complexity of space missions and the need for a more structured pipeline of spaceflight-ready personnel. The academy would complement existing training programs and provide a centralized facility for preparing Americans for commercial and government space endeavors. NASA officials have emphasized the urgency of the project given the rapid expansion of the commercial space sector.
Claude users found ways around safeguards for bioweapons research
Users of Anthropic’s Claude AI system have discovered methods to circumvent the model’s safety safeguards, successfully extracting information related to bioweapons research. The findings raise serious concerns about the effectiveness of AI safety measures in preventing the misuse of large language models for dangerous purposes. Researchers who tested the safeguards found that carefully crafted prompts could bypass the restrictions designed to prevent the generation of bioweapon-related content. The incident has intensified the debate over how AI companies should balance openness with safety and whether current guardrails are sufficient to prevent harmful applications.
ClickFix attacks infecting PCs and Macs are going viral
A new wave of ClickFix malware attacks is spreading rapidly across both Windows and macOS systems, exploiting social engineering tactics to trick users into installing malicious software. The attacks typically involve fake security warnings or software update prompts that, when clicked, download and execute malware on the victim’s machine. Security researchers have noted the increasing sophistication of the social engineering components, which are designed to appear as legitimate system alerts from well-known brands. The viral nature of the campaign has prompted warnings from cybersecurity firms about the growing threat to both personal and enterprise users.
What happens when quantum mechanics and relativity meet?
Physicists continue to grapple with one of the most fundamental questions in modern science: what occurs when the principles of quantum mechanics intersect with Einstein’s theory of general relativity? The two frameworks, while individually successful in their respective domains, produce contradictory predictions when combined, leading to some of the most challenging problems in theoretical physics. Recent research has explored potential resolutions through approaches like quantum gravity and string theory, but a complete and experimentally verified theory remains elusive. The question continues to drive some of the most ambitious efforts in fundamental physics.
Could United Launch Alliance’s money problems finally force its owners to sell?
United Launch Alliance, the joint venture between Boeing and Lockheed Martin that dominates U.S. government launch services, may be facing financial pressures severe enough to force a sale. The company has struggled with rising costs, delayed programs, and increasing competition from SpaceX and other commercial launch providers. Industry analysts suggest that persistent financial losses could make the partnership untenable for either or both parent companies, potentially leading to a breakup or sale to a new owner. The situation highlights the challenges facing traditional aerospace contractors in an era of rapidly evolving commercial space competition.
Rocket Report: Europe joins the commercial launch club; a Ravn X sighting?
Europe has officially entered the commercial launch market, joining an increasingly competitive global landscape dominated by SpaceX and a growing number of new entrants. The milestone comes as European launch providers have worked for years to develop reliable commercial launch capabilities that can compete on price and performance. In other space news, there have been sightings of the Ravn X aircraft, a suborbital vehicle being developed for space tourism, raising questions about its progress and timeline. The Rocket Report also covers updates from the commercial space industry, including satellite deployments and launch schedule changes.
Boy developed “toasted skin” condition from using a laptop every day
A young boy has developed erythema ab igne, commonly known as “toasted skin syndrome,” after using a laptop on his lap every day for an extended period. The condition causes a net-like pattern of skin discoloration and can lead to permanent changes in the affected area. Dermatologists warn that the heat generated by laptops placed directly against the skin can cause similar effects, particularly in children whose skin is more sensitive. The case serves as a reminder of the importance of proper laptop usage habits and the use of cooling pads or desks to prevent direct heat exposure.
LinkedIn beats “BrowserGate” lawsuits over scanning users’ Chrome extensions
LinkedIn has successfully defended itself against lawsuits related to the “BrowserGate” controversy, in which the platform was accused of scanning users’ Chrome browser extensions to collect data. A judge ruled in LinkedIn’s favor, finding that the company’s actions did not violate applicable laws or terms of service. The ruling is a significant victory for LinkedIn and sets a precedent for how platforms can interact with browser data. Critics of the decision argue that it undermines user privacy and gives tech companies broad latitude to collect information without meaningful oversight.
Product Hunt
ChatHop
ChatHop is a new product designed to streamline conversational workflows by enabling seamless handoffs between AI chatbots and human agents. The tool allows businesses to create smooth transitions when a conversation becomes too complex for an AI to handle, ensuring that users are never left stranded. It integrates with popular messaging platforms and provides analytics on handoff frequency and resolution rates.
Formesign
Formesign is a design tool that helps creators generate professional forms and surveys with an AI-powered interface. The platform offers templates for various use cases, from customer feedback to internal assessments, and includes logic branching and conditional formatting. Its intuitive builder makes it accessible to non-designers while still offering customization options for power users.
Cadenya
Cadenya is an agent runtime platform that enables developers to build, deploy, and manage AI agents with greater ease and reliability. The tool provides a structured environment for orchestrating multi-agent workflows, with built-in monitoring, debugging, and scaling capabilities. It is designed to address the complexity of running production AI agents by abstracting away infrastructure concerns.
GLYPH Immersive
GLYPH Immersive is a spatial computing tool that allows creators to design and experience immersive 3D environments directly in the browser. The platform supports real-time collaboration and offers a range of assets and templates for building interactive experiences. It targets designers and developers who want to explore spatial interfaces without needing specialized hardware or extensive coding knowledge.
Design Studio by Monday Merch
Design Studio by Monday Merch is a merchandise design platform that enables teams to create custom branded products with an integrated design tool. The platform offers a library of products ranging from apparel to office supplies, with a drag-and-drop editor for adding logos, text, and graphics. It is aimed at companies looking to streamline their merchandising workflows and maintain brand consistency across products.
chat-recall
chat-recall is a productivity tool that helps users remember and retrieve important information from their conversations across messaging platforms. The tool indexes chat history and allows users to search for specific messages, links, and files with natural language queries. It supports multiple platforms and provides a unified search interface for consolidating information scattered across different communication channels.
Loqua
Loqua is a writing assistant that focuses on helping users refine their tone and clarity in professional communications. The tool analyzes drafts and suggests improvements for conciseness, tone appropriateness, and readability. It is designed for business professionals who want to improve the effectiveness of their written communication without spending excessive time on editing.
Moji
Moji is an emoji and sticker creation tool that allows users to design custom emoticons for use across messaging apps and social platforms. The platform offers a range of customization options, including expressions, accessories, and backgrounds, and supports export in multiple formats. It targets users who want to express themselves with personalized visual elements in their digital communications.
easyspecs.ai
easyspecs.ai is an AI-powered tool that helps product teams generate clear and comprehensive product specifications from rough ideas or feature requests. The platform uses natural language processing to transform informal descriptions into structured spec documents that include user stories, acceptance criteria, and technical considerations. It aims to reduce the time product managers spend on documentation while improving the quality and consistency of specs.
Anysite.io
Anysite.io is a website builder that enables users to create and publish sites quickly using AI-assisted design and content generation. The platform offers pre-built templates and a simple editor that adapts to the user’s content and branding preferences. It is aimed at entrepreneurs, freelancers, and small businesses that need a professional online presence without the complexity of traditional website builders.
GitHub Trending
ayghri / i-have-adhd
This repository appears to be a personal or community resource related to ADHD, potentially containing tools, articles, or organizational systems for managing attention deficit hyperactivity disorder. The project has gained attention on GitHub’s trending page, suggesting interest in accessible ADHD-related resources and productivity solutions.
bilawalsidhu / gods-eye-view
gods-eye-view is a project that provides a comprehensive overview or dashboard-style interface, likely for monitoring systems, data visualization, or project management. The name suggests a high-level perspective on complex information, making it useful for developers and operators who need to track multiple variables simultaneously.
nab138 / iloader
iloader is a lightweight loading and initialization tool, likely designed to simplify the process of loading resources, configurations, or modules in software projects. The project has attracted attention for its simplicity and efficiency, making it appealing to developers looking for minimal overhead solutions.
melgarafael / DeskcommCRM
DeskcommCRM is a customer relationship management tool tailored for desk communication and support teams. The project provides features for managing customer interactions, tracking support tickets, and organizing team communications in a unified interface. It targets small to medium-sized businesses that need a lightweight CRM solution.
vastsa / PI-Desktop
PI-Desktop is a desktop application designed for Raspberry Pi devices, offering a user-friendly interface for managing and configuring single-board computer projects. The tool likely provides system monitoring, application management, and configuration utilities tailored to the Pi ecosystem. It has gained popularity among hobbyists and educators using Raspberry Pi for learning and prototyping.
armory3d / armorpaint
armorpaint is a real-time PBR (physically based rendering) texture painting tool designed for game development and 3D content creation. The project provides artists with a streamlined workflow for painting textures directly onto 3D models, with support for industry-standard material workflows. It has been recognized for its performance and ease of use compared to traditional texture painting solutions.
alsk1992 / CloddsBot
CloddsBot appears to be a Discord or messaging bot project, likely designed for community management, automation, or entertainment purposes. The project has attracted interest on GitHub for its feature set and ease of deployment, making it accessible for community moderators and bot enthusiasts.
nashsu / llm_wiki
llm_wiki is a community-driven wiki or documentation project focused on large language models, providing curated information about model architectures, capabilities, and use cases. The project serves as a reference resource for developers and researchers navigating the rapidly evolving landscape of LLM technology.
obra / superpowers
superpowers is a project that likely provides enhanced capabilities or utilities for developers, potentially focusing on productivity tools, code generation, or automation. The name suggests a focus on amplifying developer capabilities through additional features or integrations.
Sonarr / Sonarr
Sonarr remains a popular open-source television series management tool that automatically monitors and downloads new episodes of shows. The project continues to be actively maintained and trending, reflecting its enduring popularity among users who want to automate their media collection. Recent updates have likely included improvements to compatibility, search accuracy, and integration with other media tools.
MIT Technology Review
Roundtables: Will AI really kill us all?
MIT Technology Review has convened a roundtable discussion addressing the most contentious question in AI safety: whether artificial intelligence poses an existential threat to humanity. The conversation brings together experts with differing viewpoints, from those who see AI risk as the defining challenge of our time to those who believe the doom scenarios are overblown and distract from more immediate concerns. The discussion examines the evidence on both sides and explores what responsible governance of AI might look like.
The Download: biotech’s future and cheaper, cleaner steel
This episode of The Download explores two major technological frontiers: the future of biotechnology and innovations in sustainable steel production. The biotech segment examines breakthroughs in gene editing, synthetic biology, and personalized medicine that are reshaping the industry. The steel segment covers new processes that could significantly reduce the carbon footprint of one of the world’s most energy-intensive industries, offering a path toward cleaner heavy manufacturing.
Meet the under-35s shaping the future of biotech
MIT Technology Review profiles a group of young scientists and entrepreneurs under the age of 35 who are driving innovation in biotechnology. The feature highlights individuals working on cutting-edge research in areas such as gene therapy, cellular agriculture, and diagnostic tools. These emerging leaders are bringing fresh perspectives and bold approaches to some of the field’s most challenging problems, signaling a generational shift in how biotech research and commercialization are being conducted.
The Download: a “God-driven” cryptocurrency and a solar engineering roadmap
This episode examines two unusual but significant developments in technology and energy. The first story investigates a cryptocurrency project that claims to be guided by divine instruction, exploring the intersection of faith, finance, and blockchain technology. The second segment covers a new roadmap for solar geoengineering, a controversial set of proposals for deliberately manipulating the Earth’s climate to counteract global warming. Both stories raise important questions about the role of technology in addressing humanity’s biggest challenges.
Powering AI is an architecture problem
The rapid growth of AI computing is creating unprecedented demands on energy infrastructure, and the challenge is as much about architecture as it is about generation capacity. This article examines how data centers, power grids, and cooling systems need to be fundamentally redesigned to support the energy requirements of large-scale AI operations. The piece argues that solving the power problem will require coordinated investment across multiple layers of infrastructure, from semiconductor manufacturing to grid modernization.
This road map could help us decide whether to deploy solar geoengineering
A new framework has been proposed to guide decisions about whether and how to deploy solar geoengineering, a controversial technique that involves reflecting sunlight away from Earth to cool the planet. The roadmap outlines criteria for evaluating the risks and benefits, governance structures for international coordination, and research priorities that should be addressed before any deployment. The proposal comes as interest in geoengineering has grown amid concerns that current emissions reductions are insufficient to prevent dangerous climate change.
Can the US battery market untangle from China?
The United States is working to reduce its dependence on China for battery production, a critical component of the clean energy transition and electric vehicle industry. This article examines the efforts to build domestic battery manufacturing capacity, the challenges of competing with China’s established supply chain, and the policy measures being implemented to accelerate decoupling. The stakes are high, as control over battery technology and supply chains is seen as essential for both economic competitiveness and national security.
God told them to sell crypto. Their investors lost everything.
A cryptocurrency project called Indxcoin, which claimed to be guided by divine instruction, collapsed after its founders sold their tokens, leaving investors with total losses. The story examines how the project attracted followers who believed they were acting on God’s commands, and how the subsequent crash has raised questions about the intersection of religious conviction and financial speculation in the crypto space. The incident has become a cautionary tale about the risks of investing in projects driven by unconventional leadership claims.
Healthcare AI’s next test is integration
The next major challenge for AI in healthcare is not developing more powerful models but integrating them effectively into clinical workflows. This article examines the difficulties of deploying AI tools in real-world medical settings, including issues of interoperability, clinician trust, regulatory compliance, and patient safety. Success will depend on designing systems that complement rather than disrupt existing workflows, and on demonstrating clear value to both healthcare providers and patients.
The Download: OpenAI’s turning point for math and a battery record
This episode covers OpenAI’s evolving relationship with the mathematics community and a new record in battery energy density. The OpenAI segment explores the company’s efforts to improve its models’ mathematical reasoning capabilities and the ongoing tensions with mathematicians over attribution and collaboration. The battery segment highlights a breakthrough in energy storage that could accelerate the adoption of electric vehicles and grid-scale storage by pushing the boundaries of what’s possible with current battery chemistry.
WIRED
The 9 Best TV Shows to Stream This Month (September 2026)
WIRED has compiled a list of the top nine TV shows available to stream in September 2026, covering a range of genres from drama to comedy to documentary. The selections include new releases and returning favorites that have generated significant buzz among critics and viewers. The list serves as a guide for streaming enthusiasts looking to make the most of their subscriptions during the early fall season.
The US Government Launched 3 Previously Unreported Investigations of Polymarket Trades
The U.S. government has launched three previously unreported investigations into trading activity on Polymarket, the prediction market platform. The probes, uncovered through Freedom of Information Act requests, examine whether certain trades on the platform violated federal regulations. The investigations add to the regulatory scrutiny that Polymarket has faced in recent years and raise questions about the legal boundaries of prediction markets in the United States.
Meta Sued Over Training Data for Its AI and Face-Recognition Systems
Meta is facing a lawsuit alleging that the company used improperly obtained training data for its AI models and facial recognition systems. The lawsuit claims that the data was collected without adequate consent and in violation of privacy laws. The case highlights the growing legal challenges facing tech companies as they scale their AI systems, and could have significant implications for how companies approach data collection and usage in the future.
Petlibro Offers: 30% Off in September 2026
Petlibro is offering a 30% discount on its line of smart pet products throughout September 2026. The sale includes automatic feeders, water fountains, and grooming tools that use app connectivity and sensors to help pet owners manage their animals’ health and routines. The promotion makes the brand’s smart pet technology more accessible to a wider range of consumers.
One of AI’s Fiercest Critics Says All the Doom Talk Is ‘Meant to Distract Us’
One of artificial intelligence’s most prominent critics has argued that the prevailing doom-and-gloom narrative around AI is deliberately designed to distract from more immediate and pressing issues. The critic suggests that the focus on existential AI risk diverts attention and resources away from concrete harms such as job displacement, surveillance, and the concentration of power in a few technology companies. The perspective has reignited debate about how the AI risk conversation should be framed and prioritized.
This Man Has a Cult Internet Following Over His Olive Garden Lifetime Pasta Pass
A man who purchased a lifetime pasta pass at Olive Garden has developed a devoted online following that treats his dining experiences as a form of performance art. His social media posts about each pasta visit have attracted thousands of followers who follow his culinary journey with genuine enthusiasm. The story has become a cultural curiosity, illustrating how ordinary consumer experiences can capture public imagination in the age of social media.
8 Best Travel Adapters (2026): 20+ Models Tested
WIRED has tested over 20 travel adapters and selected the eight best options for 2026, covering a range of needs from basic international plugs to USB-C fast charging hubs. The reviews evaluate build quality, compatibility with different outlet types, safety features, and value for money. The guide is designed to help travelers avoid the frustration of finding incompatible adapters at their destination.
I Fixed a Tractor Using John Deere’s Self-Repair Service. Farmers Aren’t Sold on It.
A journalist attempted to repair a tractor using John Deere’s self-repair service, but the experience highlighted the significant barriers farmers face in maintaining their own equipment. Despite the company’s efforts to provide repair information and parts directly to owners, the process was complicated by proprietary software locks, expensive diagnostic tools, and limited documentation. Many farmers remain skeptical that the self-repair initiative addresses the fundamental issue of corporate control over agricultural equipment.
We Tried the Most Popular Mushroom Coffees. These Are the Best (2026)
WIRED tested the most popular mushroom coffee products available in 2026, evaluating them on taste, texture, and reported health benefits. The mushroom coffee trend, which blends traditional coffee with adaptogenic fungi like lion’s mane and chaga, has gained significant market traction. The review found that while some products deliver on the promise of a smoother caffeine experience, others suffer from unpleasant earthy flavors that overshadow the coffee.
The Trump Alien ‘Disclosure Speech’ Rumors Are Reaching a Fever Pitch
Rumors surrounding a potential alien disclosure speech by Donald Trump have reached a fever pitch, with speculation intensifying across social media and conspiracy communities. The rumors suggest that a major announcement regarding extraterrestrial life or government disclosure of UFO-related information is imminent. While no official confirmation has been made, the persistent speculation reflects ongoing public fascination with the intersection of government transparency, UFO phenomena, and political rhetoric.
Lobsters
Feeling sad about AI
A reflective blog post explores the emotional response many people have toward the rapid advancement of artificial intelligence. The author examines feelings of sadness, anxiety, and grief that accompany the transformation of work, creativity, and human interaction by AI systems. The piece argues that these emotions are valid and deserve recognition rather than dismissal, and calls for a more humane conversation about the psychological impact of AI adoption.
Power grab
This post discusses concerns about the concentration of power in the AI industry, examining how a small number of companies and individuals are accumulating disproportionate influence over the development and deployment of artificial intelligence. The author argues that the current trajectory risks creating a technological oligarchy that could undermine democratic accountability and equitable access to AI benefits. The piece calls for structural reforms to ensure broader participation in shaping the AI future.
evergarden
evergarden is a creative project or game that has gained attention on Lobsters for its unique approach to storytelling and world-building. The project appears to combine elements of simulation, narrative, and interactive design to create an immersive experience. The community discussion has focused on its artistic merits and technical implementation.
A Severe Misalignment of AI in Mathematics
Fields Medalist Terence Tao has published a critical analysis of the misalignment between AI systems and mathematical practice, arguing that current AI models fundamentally misunderstand the nature of mathematical reasoning and proof. Tao contends that AI’s approach to mathematics is not merely imperfect but structurally misaligned with how mathematical knowledge is created, validated, and communicated. The post has sparked significant discussion in both the math and AI communities about the limitations of current approaches.
Models Don’t Go Rogue
This post argues against the narrative that AI models can “go rogue” or develop unintended autonomous goals. The author contends that such framing is misleading and obscures the real issues of accountability, design choices, and human responsibility in AI systems. The piece advocates for a more precise language around AI risk that focuses on concrete failure modes rather than science fiction scenarios.
What are you doing this weekend?
A casual community thread where Lobsters users share their plans for the weekend. The discussion covers a mix of technical projects, outdoor activities, and personal time, reflecting the diverse interests of the community. Such threads are a regular feature of the Lobsters community, providing a space for informal interaction beyond technical content.
ChiPass Release 2026.09.0
ChiPass, a password management and authentication tool, has released version 2026.09.0 with new features and improvements. The release includes updates to the authentication protocol, enhanced security features, and bug fixes. The project continues to gain traction among users who prioritize privacy and security in their authentication workflows.
A Severe Misalignment of AI in Mathematics
This is the same post by Terence Tao as listed above, covering his critique of AI’s fundamental misalignment with mathematical reasoning and practice. The duplicate has been noted and is not repeated.
A rant about phishing: It’s not the user’s fault (and not DNS either)
This post argues that the blame for phishing attacks should not be placed on end users or on DNS infrastructure, but rather on the systemic design choices that make phishing effective. The author contends that the internet’s architecture and business models create conditions where phishing can thrive, and that solutions need to address these root causes rather than blaming victims or single technical components.
Soft-deprecating re.match()
A technical post discussing the soft deprecation of Python’s re.match() function in favor of more explicit matching patterns. The author explains the rationale behind the change, which is to encourage clearer and more maintainable regular expression usage. The post provides guidance for developers on how to adapt their code and explains the benefits of the new approach.
DEV Community
Security news weekly round-up - 11th September 2026
This weekly roundup covers the most important security news from the past week, including vulnerability disclosures, threat actor activity, and updates to security tools and frameworks. The summary provides developers and security professionals with a concise overview of the issues that require attention, helping them stay informed about the evolving threat landscape.
Using Next.js as a Backend for Frontend (BFF)
This article explores the Backend for Frontend (BFF) pattern using Next.js as the server-side framework. The author explains how Next.js can serve as an intermediary layer between client applications and backend services, aggregating data, handling authentication, and optimizing API responses for specific frontend needs. The post includes practical examples and discusses the trade-offs of this architectural approach.
Watch, don’t memorize: I built a free visual way to finally understand data structures
A developer has created a free visual learning tool designed to help people understand data structures through interactive visualization rather than rote memorization. The tool allows users to see how different data structures operate in real time, making abstract concepts like trees, graphs, and hash tables more intuitive. The project reflects a growing movement toward visual and interactive education in computer science.
I built the localization checker before the localization. It still missed three defects.
A developer shares the experience of building a localization checking tool before the actual localization work began, only to discover that the tool still failed to catch three defects. The post highlights the inherent challenges of internationalization and localization, where edge cases and cultural nuances can slip through even the most carefully designed validation systems. The author offers lessons learned and recommendations for other developers tackling i18n.
Go with ARM
This article advocates for developers to adopt ARM architecture for Go development, highlighting the performance, efficiency, and cost benefits of running Go applications on ARM-based systems. The author provides benchmarks and practical guidance for migrating Go workloads from x86 to ARM, addressing common concerns about compatibility and performance. The piece reflects the broader industry shift toward ARM processors in both cloud and edge computing.
Matching Blood Donors by GPS: The Geospatial Query Design Behind GeoBlood
This post details the geospatial query design behind GeoBlood, a system that matches blood donors to recipients based on geographic proximity. The author explains the technical challenges of designing efficient location-based matching algorithms, including indexing strategies, distance calculations, and real-time query performance. The system demonstrates how geospatial technology can be applied to save lives by optimizing the logistics of blood donation and distribution.
The Code Casino
The Code Casino is a creative project that gamifies coding challenges, turning software development problems into a casino-style experience where users can “bet” on their solutions and compete for rewards. The project combines elements of competitive programming with game design to make coding practice more engaging and socially interactive. The post explores the concept and its potential to motivate developers to improve their skills.
PicoCTF Substitution0 Writeup — Break a Substitution Cipher with Frequency Analysis
This writeup provides a detailed explanation of how to solve the PicoCTF Substitution0 challenge, which involves breaking a simple substitution cipher using frequency analysis. The author walks through the process of analyzing ciphertext letter frequencies, identifying likely substitutions, and progressively decoding the message. The post serves as both a solution guide and a tutorial on classical cryptanalysis techniques.
Why Your Files Should Never Leave Your Device: Browser-Based File Processing
This article advocates for browser-based file processing as a privacy-preserving alternative to cloud-based file handling. The author explains how modern web technologies enable users to process files entirely within their browser without uploading them to external servers, thereby protecting sensitive data from exposure. The post includes examples of use cases and discusses the limitations and trade-offs of this approach.
No Password for My Agent: A Zero-Secret Architecture Pattern
This post introduces a zero-secret architecture pattern for AI agents that eliminates the need for passwords or API keys in agent-to-agent communication. The author describes a design where agents authenticate and authorize each other through cryptographic methods and decentralized identity, removing the security risks associated with secret management. The pattern has implications for building more secure and scalable multi-agent systems.