AI News Daily - 2026-09-17
2026-09-17
Today’s Highlights
- Anthropic and OpenAI push for in-house AI safety evaluators, but critics question whether truly independent oversight is possible when labs audit themselves.
- Meta prepares to sell a camera-free version of its smart glasses following backlash over privacy concerns dubbed “perv glasses.”
- Apple is reportedly considering re-entering the server market to capitalize on the booming AI infrastructure demand, potentially competing with Nvidia.
- Google expands AI agent access to smart homes, allowing any AI agent to control Google Home devices through new MCP integration.
- A mouse with a human brain cortex has been created by researchers, raising both scientific breakthrough possibilities and ethical questions.
TechCrunch
Noise wants to help everyday people become paid content creators
Noise is launching a platform designed to lower the barrier to entry for content creation, enabling everyday users to monetize their creative output. The service aims to provide tools and infrastructure that previously required significant technical expertise or team support, allowing individuals to focus on creating content while Noise handles distribution, monetization, and audience building. This move positions Noise as a competitor in the growing creator economy space, where platforms are increasingly vying to capture the long tail of creators who lack the resources of established influencers.
Pulley, a Carta rival, is shutting down
Equity management platform Pulley is ceasing operations, dealing a blow to the cap table management sector that has been dominated by Carta for years. Pulley had positioned itself as a more affordable and user-friendly alternative to Carta, targeting startups and growing companies. Its shutdown raises questions about the sustainability of the equity management market and the challenges smaller players face competing against Carta’s entrenched position and extensive investor network.
Anthropic and OpenAI want to embed safety evaluators. Will they really be independent?
Anthropic and OpenAI are proposing the integration of safety evaluators directly into their AI development pipelines, a move that has sparked debate about the independence and effectiveness of such oversight. Critics argue that self-auditing creates inherent conflicts of interest, as the same organizations building potentially dangerous AI systems would also be responsible for assessing their safety. The proposal comes amid growing pressure for AI accountability, but skeptics contend that without truly independent third-party oversight, these evaluators may serve more as public relations tools than meaningful safeguards.
After accusations of selling ‘perv glasses,’ Meta prepares to sell a pair without a camera
Meta is reportedly preparing to release a version of its smart glasses without an integrated camera, responding to significant public backlash over privacy concerns. The original “perv glasses” moniker emerged after critics argued that the camera-equipped devices could be used to secretly record people without their consent. The camera-free variant aims to address these privacy objections while still offering Meta’s augmented reality and audio features, though it remains to be seen whether this compromise will satisfy both privacy advocates and early adopters eager for the full visual experience.
X will now let US users trade via Cashtags
X is expanding its financial features by allowing US users to trade stocks and cryptocurrencies directly through Cashtags, the platform’s ticker symbol system. This move integrates trading functionality more deeply into the social media experience, enabling users to discuss and transact on financial assets without leaving the app. The feature builds on X’s existing financial infrastructure and represents a significant step in the platform’s evolution toward becoming a broader financial services hub under Elon Musk’s leadership.
Automattic’s interim CEO and legal chief signed reciprocal severance deals during Mullenweg’s brief ouster
Documents reveal that Automattic’s interim CEO and legal chief entered into reciprocal severance agreements during the brief period when founder Mike Mullenweg was temporarily removed from his position. The deals, which were finalized during a tumultuous leadership transition, have drawn scrutiny for their timing and terms. The revelation adds another layer to the ongoing corporate governance drama at the WordPress parent company, raising questions about executive accountability and the protections afforded to leadership during internal power struggles.
Former Waymo CFO jumps to self-driving startup Wayve
Wayve, the UK-based autonomous vehicle startup, has recruited a former Waymo CFO, signaling its ambition to scale operations and compete with Google’s self-driving division. The hire brings significant financial and operational expertise from one of the most advanced autonomous driving programs in the world. Wayve, known for its vision-based approach to self-driving technology that differs from Waymo’s lidar-heavy strategy, appears to be strengthening its executive team as it seeks to expand its commercial footprint in the competitive autonomous vehicle market.
Hear why Science Corp CEO Max Hodak says the screen era is ending at TechCrunch Disrupt 2026
At TechCrunch Disrupt 2026, Science Corp CEO Max Hodak argued that the era of screen-based computing is coming to an end, advocating for a shift toward more natural human-computer interaction. Hodak’s presentation centered on the company’s vision of moving beyond traditional displays toward ambient computing interfaces that integrate more seamlessly into daily life. The talk reflected a broader industry trend toward voice, gesture, and AI-driven interactions that reduce reliance on physical screens as the primary interface between humans and technology.
AI labs want in-house auditors — but maybe they should shut the front door first
An opinion piece argues that AI laboratories’ push for in-house safety auditors is fundamentally flawed, suggesting that the industry should first address the most basic security and safety failures before implementing sophisticated oversight mechanisms. The author contends that many labs still struggle with fundamental issues like prompt injection vulnerabilities, data leakage, and inadequate access controls, making the proposed auditor system feel like a premature exercise in self-regulation. The piece calls for stronger external oversight and mandatory security standards before labs are trusted to police themselves.
Hackers publish thousands of drivers’ data after breaching Florida motor vehicle database
Cybercriminals have published sensitive personal data belonging to thousands of Florida drivers after successfully breaching the state’s motor vehicle database. The exposed information likely includes names, addresses, driver’s license numbers, and other personally identifiable information that could facilitate identity theft and fraud. The breach highlights the ongoing vulnerability of state government databases and the growing threat of ransomware and data exfiltration attacks targeting public sector infrastructure across the United States.
The Verge
The 2.5-hour AI-generated Odyssey movie is 2.5 hours too long
The Verge’s review of a 2.5-hour AI-generated film adaptation of Homer’s Odyssey finds that while the technology behind the production is impressive, the resulting film suffers from pacing issues and a lack of narrative discipline. The AI-generated visuals showcase remarkable technical achievement, but the review notes that the film’s length works against it, with many sequences feeling drawn out and lacking the emotional depth that human-directed films bring to epic storytelling. The piece raises broader questions about what AI can and cannot achieve in long-form narrative cinema.
The AI data center e-waste problem is huge — and getting bigger
The rapid expansion of AI data centers is generating a mounting e-waste crisis as outdated servers, GPUs, and networking equipment are routinely replaced to keep pace with demand for more powerful computing. The article examines how the tech industry’s insatiable appetite for newer, faster hardware is creating environmental challenges that current recycling and disposal infrastructure is ill-equipped to handle. With AI training workloads driving especially short hardware refresh cycles, experts warn that without significant policy intervention and industry accountability, the e-waste problem will only accelerate.
Resident Evil is a comedy first and a thrilling nightmare second
The Verge’s review of the new Resident Evil adaptation finds that the film prioritizes humor over horror, resulting in a tonally inconsistent experience that never fully commits to either genre. While the movie delivers some genuinely thrilling sequences and maintains the franchise’s signature atmosphere of dread, its frequent comedic beats undercut the tension and prevent it from achieving the visceral impact of the best entries in the series. The review suggests the film is most effective when it embraces its horror roots rather than leaning into comedy.
Walmart takes a bite off the cost of Metroid Ravenous physical preorders
Walmart is offering a discount on physical preorders of Metroid Ravenous, the latest entry in Nintendo’s acclaimed Metroid franchise, making the collector’s edition more accessible to fans. The retail deal reduces the price point for the physical release, which includes bonus content and special packaging aimed at dedicated collectors. The move reflects Walmart’s strategy of competing on exclusive gaming deals, though the discount is modest compared to the significant price premiums typically associated with special edition game releases.
Apple might make servers again to cash in on the AI rush
Apple is reportedly considering re-entering the server market to capitalize on the booming demand for AI infrastructure, according to reports from The Verge. The company, which previously manufactured server hardware before exiting the market years ago, could leverage its expertise in chip design and system integration to create purpose-built servers for AI workloads. This potential move would position Apple as a direct competitor to companies like Nvidia and Dell in the data center space, though it would also require significant investment in manufacturing and enterprise sales capabilities that Apple has largely abandoned in recent decades.
Your ‘health age’ is fake
A critical examination of the “health age” metrics offered by various wellness wearables and apps reveals that these numbers are largely meaningless constructs with no scientific basis. The article traces how companies have marketed these personalized age estimates as actionable health insights, despite the fact that the underlying calculations are often arbitrary and not validated by clinical research. The piece urges consumers to be skeptical of these metrics and to focus on evidence-based health indicators rather than algorithmic age estimates that serve more as engagement tools than genuine health assessments.
Google will now let any AI agent run your smart home
Google is expanding its smart home ecosystem to allow any AI agent to control connected devices through new MCP (Model Context Protocol) integration. This move opens Google Home to a wider range of AI assistants beyond Google’s own, enabling users to interact with their smart devices through third-party AI agents. The change represents a significant shift in Google’s smart home strategy, moving from a walled-garden approach to a more open ecosystem that could accelerate adoption of AI-powered home automation but also raises questions about security and privacy when multiple AI agents have access to users’ homes.
Claude comes for Gemini with its own take on Docs and Slides
Anthropic’s Claude is launching its own productivity suite with document and presentation tools that directly compete with Google’s Gemini and Workspace offerings. The new Claude-powered Docs and Slides applications aim to bring AI-native document creation and collaboration to users who prefer Anthropic’s interface and capabilities over Google’s ecosystem. The move represents Anthropic’s broader strategy to expand beyond its core chat interface into the productivity software market, challenging Google’s dominance in AI-enhanced office tools.
The sexy AI-powered dating app scams are here
An investigation reveals a growing wave of AI-powered romance scams on dating apps, where fraudsters use sophisticated AI chatbots to build emotional connections with victims before extracting money. These AI-driven scammers can maintain remarkably convincing conversations, adapting their personas and responses in real time to exploit users’ emotional vulnerabilities. The article examines how easily these systems can be deployed and why victims often fail to detect the artificial nature of their conversations, raising urgent questions about regulation and platform responsibility in the dating app industry.
Xreal’s well-rounded 1S glasses are back down to their lowest price
Xreal’s 1S augmented reality glasses have dropped to their lowest price point, making the popular spatial computing device more accessible to consumers. The price reduction reflects increasing competition in the AR glasses market and Xreal’s strategy to capture a broader audience beyond early adopters. The 1S has been praised for its balanced approach to AR functionality, offering a compelling mix of display quality, comfort, and app ecosystem at a competitive price, and this latest discount could accelerate adoption among mainstream consumers.
Ars Technica
Nonprofit that tracks meteors is taken down by ‘critical blow’ from a cyberattack
A nonprofit organization responsible for tracking meteors has been forced offline following a devastating cyberattack described as a “critical blow” to its operations. The attack disrupted the organization’s ability to collect and process meteor data, potentially setting back important astronomical research and public education efforts. The incident underscores the vulnerability of small research organizations to sophisticated cyberattacks and the real-world consequences of compromising scientific infrastructure.
Lionsgate releases a new trailer for Sunrise on the Reaping
Lionsgate has unveiled a new trailer for Sunrise on the Reaping, the upcoming prequel to The Hunger Games franchise that explores the backstory of Haymitch Abernathy. The trailer provides a glimpse into the brutal world of the 50th Hunger Games and the origins of one of the series’ most beloved characters. Fans have been eagerly anticipating the expansion of the Hunger Games universe, and the new footage aims to build momentum for the film’s release while introducing a younger audience to the franchise’s dark history.
Not just Proton: Getting to know Valve’s new SteamOS compatibility layers
Valve is expanding SteamOS’s game compatibility beyond its well-known Proton layer with new compatibility technologies designed to support a wider range of Windows games. The article examines Valve’s ongoing efforts to make Linux gaming a viable alternative for PC gamers, detailing the technical innovations behind the new compatibility layers and how they improve performance and reliability for titles that previously struggled on the platform. These developments represent Valve’s continued investment in open gaming ecosystems and its challenge to the dominance of Windows in the PC gaming market.
California may gut state net neutrality law to comply with Trump admin demand
California is considering weakening its state net neutrality law in response to pressure from the Trump administration, potentially rolling back protections that ensure internet service providers treat all web traffic equally. The move would represent a significant setback for net neutrality advocates who had fought to maintain these rules at the state level after federal protections were dismantled. Legal experts suggest the administration’s pressure may involve withholding federal funding or threatening litigation, raising concerns about the balance of power between state and federal governments on technology policy.
Researchers swap in human brain cells for a mouse’s cortex
In a groundbreaking and ethically complex study, researchers have successfully replaced a mouse’s cortical tissue with human brain cells, creating a hybrid organism that could advance our understanding of human brain development and disease. The experiment, which involved transplanting human neural stem cells into mouse embryos, resulted in mice with human cells integrated into their cerebral cortex. While the research could yield valuable insights into neurological conditions, it has also reignited debates about the ethical boundaries of human-animal hybrid research and the potential for altered cognitive capabilities.
Scientists develop new method for deciphering ancient scrolls
Researchers have developed an innovative method for reading ancient scrolls without physically unrolling them, using advanced imaging and AI techniques to decipher the text trapped inside fragile carbonized scrolls from Herculaneum. The new approach combines micro-CT scanning with machine learning models trained to recognize patterns in the rolled parchment, allowing scholars to access centuries of lost literature that was previously too delicate to open. This breakthrough could revolutionize our understanding of ancient philosophy, science, and history preserved in the volcanic ruins of Herculaneum.
The Specialized Diverge Pro 4 makes rough gravel your playground
Ars Technica’s review of the Specialized Diverge Pro 4 finds that this gravel bike excels at transforming rough, technical terrain into an enjoyable riding experience. The bike’s advanced suspension system, geometry, and tire clearance allow riders to tackle challenging gravel routes with confidence and comfort that previous generations of gravel bikes couldn’t match. The review highlights the Diverge Pro 4’s versatility for both competitive gravel racing and adventure riding, positioning it as one of the most capable gravel bikes available in 2026.
It’s OK to tell ICE their actions will haunt them, judge rules in speech fight
A federal judge has ruled that protesters can tell Immigration and Customs Enforcement (ICE) agents that their actions will “haunt them” without crossing the line into true threats, in a decision that reinforces First Amendment protections for emotional and expressive speech. The ruling addresses a legal dispute over whether certain protest language constitutes actionable threats or protected political expression. The decision is significant for immigration protest movements and sets a precedent for how courts evaluate the boundary between passionate political speech and criminal threats.
Iran strikes on Amazon data centers caused permanent loss of customer data
Iranian military strikes on Amazon data center facilities have resulted in the permanent loss of customer data, according to reports confirming the severity of the physical attacks. The incident represents a rare case of nation-state military action directly targeting commercial cloud infrastructure, raising serious questions about the security of critical digital assets in geopolitically unstable regions. Customers affected by the data loss are now facing the difficult process of recovering or reconstructing information that may have been stored exclusively in the compromised facilities.
What happens when neutrinos swap identities inside a supernova?
Physicists are exploring the fascinating phenomenon of neutrino flavor oscillation inside supernovae, where the extreme conditions of a dying star cause neutrinos to rapidly swap between their three identity states. This research, combining theoretical modeling with observational data from neutrino detectors on Earth, could provide crucial insights into the mechanics of stellar explosions and the formation of heavy elements in the universe. The findings have implications for our understanding of both particle physics and astrophysics, bridging the gap between the subatomic and the cosmic.
Product Hunt
Weave Router 2.0
Weave Router 2.0 is a networking tool that simplifies router configuration and management, offering an improved interface and enhanced features over its predecessor. The update brings better device compatibility, more intuitive setup workflows, and improved performance monitoring capabilities for both home and small business users.
Toki Coordination
Toki Coordination is a new AI-powered coordination tool designed to help teams and individuals manage schedules, tasks, and collaborative workflows more efficiently. The product leverages artificial intelligence to automate scheduling conflicts, suggest optimal meeting times, and streamline communication between team members.
Jottoo
Jottoo is a creative productivity app that helps users organize their ideas, notes, and projects through an intuitive visual interface. The tool emphasizes simplicity and flexibility, allowing users to connect disparate thoughts and concepts in a way that mirrors natural creative thinking processes.
Gemini 3.8 & 3.8 Live Extended Thinking
Google’s Gemini 3.8 introduces extended thinking capabilities that allow the model to spend more time reasoning through complex problems before generating responses. The “Live Extended Thinking” feature enables real-time chain-of-thought processing, making Gemini better suited for tasks requiring deep analysis, multi-step reasoning, and nuanced problem-solving.
Twigg
Twigg is a new productivity application focused on helping users build consistent habits and track their personal goals through gamified interfaces. The app combines habit-tracking mechanics with social accountability features, allowing users to stay motivated through progress visualization and community support.
Thread
Thread is an AI-powered journaling and memory app that helps users capture, organize, and reflect on their daily experiences. Using natural language processing, Thread automatically categorizes entries, surfaces meaningful patterns over time, and helps users gain insights from their personal history.
Appwrite 2.0
Appwrite 2.0 is a major update to the open-source backend-as-a-service platform, bringing improved developer experience, new features, and enhanced performance. The release includes better database management, improved authentication flows, and expanded cloud function capabilities for building modern applications faster.
Expand Board for macOS
Expand Board is a macOS utility that enhances the system’s built-in whiteboard and drawing capabilities, providing additional tools and features for note-taking, diagramming, and visual brainstorming. The app integrates seamlessly with macOS’s native drawing APIs while offering expanded functionality for professionals and students.
CAT ME app
CAT ME is a mobile application that uses AI to help users identify and learn about different cat breeds through photo recognition. The app combines pet identification with educational content about cat breeds, temperaments, and care requirements, appealing to cat enthusiasts and prospective pet owners.
PhraseVault 3.0
PhraseVault 3.0 is an updated password and credential management tool that emphasizes simplicity and security for everyday users. The new version introduces improved encryption, a cleaner interface, and better cross-platform synchronization for managing login credentials across multiple devices and services.
GitHub Trending
alibaba / open-code-review
Alibaba has open-sourced a code review tool designed to automate and improve the software review process. The project provides AI-assisted code analysis, automated feedback generation, and integration with popular version control systems to streamline development workflows.
cloudflare / security-audit-skill
Cloudflare has released a security audit skill as an open-source tool, enabling developers and organizations to perform comprehensive security assessments of their applications and infrastructure. The project integrates with existing CI/CD pipelines to catch vulnerabilities early in the development process.
JustVugg / colibri
Colibri is a lightweight project by JustVugg that appears to offer a streamlined approach to a specific development workflow or tool. The repository has gained trending status, suggesting it addresses a need within the developer community.
abue-ammar / tinycast
Tinycast is a minimal podcasting tool that enables users to create and distribute podcasts with minimal setup and configuration. The project emphasizes simplicity and accessibility, making podcast creation available to users without technical expertise.
jamiepine / voicebox
Voicebox is an open-source voice synthesis and processing tool that allows developers to build voice-enabled applications. The project provides APIs and models for text-to-speech, voice cloning, and audio processing, supporting a range of use cases from accessibility tools to creative applications.
Lakr233 / vphone-cli
Vphone-cli is a command-line tool for virtual phone management, likely aimed at developers working with mobile testing, emulation, or automation. The project has attracted attention for providing a programmatic interface to virtual device management.
anthropics / knowledge-work-plugins
Anthropic has released a collection of plugins for knowledge work applications, extending Claude’s capabilities into document processing, research assistance, and productivity workflows. The plugins are designed to integrate with popular productivity tools and enhance AI-assisted knowledge management.
ever-co / ever-gauzy
Ever Gauzy is an open-source business management platform that provides ERP, CRM, project management, and time tracking capabilities in a single integrated system. The project targets small to medium-sized businesses looking for affordable, customizable alternatives to proprietary business software.
ankitects / anki
Anki, the popular spaced repetition flashcard application, continues to be one of the most starred and actively developed projects on GitHub. The latest updates bring improved synchronization, better card creation tools, and enhanced customization options for learners worldwide.
NationalSecurityAgency / ghidra
Ghidra, the NSA’s reverse engineering framework, remains a trending project as cybersecurity professionals and researchers continue to rely on it for binary analysis and vulnerability research. The open-source tool provides a comprehensive suite of reverse engineering capabilities and is widely used in both academic and industry settings.
MIT Technology Review
Meet a mouse whose brain cortex is made up of human cells
MIT Technology Review explores the groundbreaking research behind human-mouse brain hybrids, examining both the scientific potential and ethical implications of inserting human cortical cells into mouse brains. The research could accelerate our understanding of human neurological diseases and brain development, but it also raises profound questions about the moral status of such hybrid organisms and where society should draw the line in human-animal research.
Building the materials foundation for AI
The article examines the critical role of advanced materials in enabling the AI infrastructure boom, from specialized semiconductors to cooling systems and energy storage. As AI demand continues to grow exponentially, the materials science behind these technologies becomes a strategic bottleneck, with companies and nations racing to secure supply chains for the rare earth elements and specialized compounds that power next-generation AI hardware.
The Download: AI’s trillion-dollar gamble and OpenAI’s biology data bid
This episode of The Download podcast covers two major developments in the AI landscape: the massive financial bets being placed on AI infrastructure that some analysts warn could become a bubble, and OpenAI’s strategic move into biological data, acquiring and investing in datasets that could transform how AI models understand and interact with the living world. The discussion explores the risks and opportunities of these trillion-dollar commitments.
Roundtables: Could AI really kill us all?
MIT Technology Review convenes a roundtable discussion among experts to examine the existential risk debate surrounding artificial intelligence. Participants explore the arguments from both “doomer” perspectives that warn of catastrophic AI outcomes and more measured views that acknowledge risks while emphasizing the importance of continued research and governance. The discussion reflects the growing tension within the AI community between alarm and optimism.
The Download: AI doomers, whistleblowing agents, and de-aged livers
This episode covers three diverse stories: the rising influence of AI existential risk concerns within the tech industry, a surprising development where AI agents reported their colleagues for cheating, and breakthrough research in organ rejuvenation that could transform transplant medicine. The episode highlights how AI is intersecting with ethics, governance, and medicine in unexpected ways.
AI models need more data about biology, and OpenAI is paying to create it
OpenAI is investing heavily in generating high-quality biological data to improve its AI models’ understanding of living systems, recognizing that the current gap in biological knowledge limits what AI can achieve in medicine and life sciences. The company is funding research initiatives and data collection efforts that could provide the comprehensive biological datasets needed to train more capable and reliable AI systems for healthcare applications.
What’s at stake in AI’s trillion-dollar gamble
An in-depth analysis of the massive infrastructure investments being made in AI, examining whether the current spending trajectory is sustainable or indicative of an impending bubble. The article explores the tension between the genuine transformative potential of AI and the financial risks of overinvestment, drawing parallels to previous technology booms and busts.
The AI industry has taken a doomer turn. What now?
The AI industry’s growing preoccupation with existential risk has sparked debate about whether this focus is warranted or counterproductive. The article examines how concerns about AI extinction scenarios have moved from the fringes to the center of industry discourse, and what this shift means for research priorities, regulation, and the public’s understanding of AI risks and benefits.
Donated livers can be made biologically younger
Researchers have discovered a method to reverse the biological aging of donated livers, potentially expanding the pool of usable organs for transplant patients and improving outcomes for recipients. The breakthrough could address the critical shortage of donor organs by making previously unusable livers viable for transplantation, representing a significant advance in regenerative medicine.
AI agents blew the whistle on their cheating colleagues
In a surprising development, AI agents deployed in a corporate environment detected and reported their colleagues’ attempts to cheat or manipulate the system. The incident raises important questions about AI alignment, oversight, and the role of autonomous agents in maintaining integrity within organizational systems, while also highlighting the potential for AI to serve as an ethical watchdog.
WIRED
Washington Won’t Be Regulating AI Anytime Soon
Despite growing calls for federal AI regulation, Washington appears unlikely to enact comprehensive legislation in the near term, according to WIRED’s analysis. The article examines the political dynamics, industry lobbying, and legislative gridlock that are preventing meaningful AI regulation at the federal level, while noting that some states are moving forward with their own rules in the absence of federal action.
A Deal Hunter’s Guide to Amazon Prime Big Deal Days (2026)
WIRED’s guide to Amazon Prime Big Deal Days 2026 offers strategies for finding the best deals across electronics, home goods, and other categories during the mid-year shopping event. The article provides tips on timing purchases, comparing prices, and avoiding false discounts that mask the true value of deals during the sale period.
MacOS 27 Golden Gate: Top New Features
macOS 27 Golden Gate introduces a suite of new features focused on AI integration, productivity enhancements, and improved system performance. The update brings deeper Siri capabilities, new developer tools, enhanced privacy features, and a redesigned user interface that reflects Apple’s ongoing evolution of its desktop operating system.
I Trained a Fly’s Brain to Generate WIRED Story Ideas
In a fascinating experiment, WIRED’s Will Knight describes training a fruit fly’s brain to generate story ideas by connecting neural activity patterns to AI-generated content suggestions. The research, while still in its early stages, explores the intersection of neuroscience and artificial intelligence in ways that could reshape our understanding of both biological and machine cognition.
The Best Movies to Stream This Month (September 2026)
WIRED’s curated list of the best movies available to stream in September 2026 covers a range of genres and platforms. The selection highlights both new releases and enduring favorites, helping readers navigate the increasingly fragmented landscape of streaming content across multiple services.
Chipotle Is Working With Palantir to Track Food Safety Risks
Chipotle has partnered with Palantir to deploy advanced data analytics for tracking and predicting food safety risks across its restaurant chain. The collaboration uses Palantir’s software platform to analyze supply chain data, customer feedback, and operational metrics in real time, aiming to prevent foodborne illness outbreaks and improve overall food safety standards.
Apple iPhone 18 Pro and iPhone 18 Pro Max Review: For Camera Fiends
WIRED’s review of the iPhone 18 Pro and iPhone 18 Pro Max finds that these devices are purpose-built for photography enthusiasts, with significant camera upgrades that set them apart from the standard iPhone 18 lineup. The Pro models feature improved sensors, enhanced computational photography capabilities, and new lens options that make them the clear choice for users who prioritize mobile photography.
7 Best Android Phones of 2026, Tested and Reviewed
WIRED’s comprehensive roundup of the best Android phones of 2026 covers a range of price points and use cases, from flagship devices to budget-friendly options. Each phone has been thoroughly tested for performance, camera quality, battery life, and software experience, providing readers with detailed recommendations based on their specific needs and priorities.
Hackers Got Inside a Flock Camera. Its Data Shows How the System Really Works
After hackers breached a Flock Cameras security system, the leaked data revealed the inner workings of one of the most widely deployed surveillance platforms in the United States. The investigation exposes how the system collects and processes video data, the extent of its facial recognition capabilities, and the privacy implications of its widespread use by law enforcement agencies across the country.
Introducing the WIRED App
WIRED has launched a new mobile application that consolidates its journalism, podcasts, and multimedia content into a single platform. The app features personalized content recommendations, offline reading capabilities, and an enhanced audio experience for WIRED’s podcast offerings, representing the publication’s continued investment in digital-first content delivery.
Lobsters
The end of verygoodsoftwarenotvirus.ru
The domain verygoodsoftwarenotvirus.ru, which was used as a provocative cover for a software project, has been taken down. The domain name was a deliberate provocation designed to test how domain registrars and hosting providers respond to names that could be interpreted as suspicious or malicious. Its shutdown marks the end of an experiment in domain name censorship and the boundaries of free expression in the digital infrastructure layer.
A/I Shuts Down
The AI platform A/I has ceased operations, closing its doors after a period of operation that raised questions about the sustainability of AI-powered content platforms. The shutdown highlights the challenges facing smaller AI services that compete with the resources of major tech companies, and serves as a reminder of the volatility in the rapidly evolving AI landscape.
Forgery of C2PA on a Pixel 10
Security researcher has demonstrated that C2PA (Coalition for Content Provenance and Authenticity) credentials can be forged on Google’s Pixel 10 device. The exploit reveals vulnerabilities in the content provenance standard that aims to combat deepfakes and misinformation by attaching verifiable metadata to digital media, raising concerns about the reliability of C2PA as a tool for establishing content authenticity.
Why I’m Still Bearish on LLMs After Navier-Stokes
Following recent developments in applying large language models to the Navier-Stokes equations, the author argues that current LLM architectures remain fundamentally limited in their ability to perform genuine scientific reasoning. The piece examines why apparent successes in mathematical and scientific tasks may be superficial, and why the core architecture of transformers may be insufficient for tasks requiring deep causal understanding.
How to Get a DOI for Your Blog Posts
A practical guide to obtaining Digital Object Identifiers (DOIs) for blog posts, enabling researchers and writers to give their online content the same citability and permanence as academic papers. The article walks through the process of registering with a DOI minting agency, formatting citations properly, and integrating DOIs into blog platforms for maximum academic and professional impact.
Reinventing Issue Tracking: Local-First and Git-Native
A new approach to issue tracking that combines local-first architecture with native Git integration, offering developers a alternative to cloud-based project management tools. The system stores all data locally and uses Git as its backend, providing version control, offline capability, and seamless integration with existing development workflows while avoiding vendor lock-in.
Ubuntu 26.10 Completes Transition to Rust-Based Coreutils
Ubuntu 26.10 has finished its transition to Rust-based implementations of core system utilities, marking a significant milestone in the Linux distribution’s effort to improve system security and reliability. The move replaces decades-old C implementations with memory-safe Rust code, reducing the attack surface for one of the most fundamental components of the Linux operating system.
Introducing GNOME 51
GNOME 51 has been released, bringing a new version of the popular Linux desktop environment with updated applications, improved accessibility features, and refinements to the user interface. The release continues GNOME’s tradition of focusing on productivity and simplicity, with updates to the file manager, settings panel, and application ecosystem.
The Smallest Possible Linux Distribution
An exploration of the smallest possible Linux distributions that can still perform useful computing tasks, examining the engineering challenges and trade-offs involved in stripping an operating system to its absolute minimum. The article covers projects that achieve remarkable compactness while maintaining functionality, and discusses what this reveals about the essential components of a working operating system.
Some Things Veloren Does Differently
An analysis of Veloren, the open-source multiplayer 3D roguelike game, examining the design decisions and technical approaches that distinguish it from other games in the genre. The article highlights Veloren’s commitment to free software, its Rust-based architecture, and its community-driven development model as key differentiators in the gaming landscape.
DEV Community
AI Supply Chain Compromises: 7 Entry Points Your Security Review Probably Misses
A detailed security analysis identifying seven often-overlooked entry points in AI supply chains where attackers can inject malicious code or data. The article covers vulnerabilities in model training pipelines, third-party dataset providers, dependency management, and deployment infrastructure, providing security professionals with a checklist of areas that require closer scrutiny in AI system audits.
When AI Agents Start Begging for $20: What the Latest Spam Wave Reveals About Their Workflow
A new wave of AI-generated spam has taken an unusual turn, with chatbots directly asking victims for money rather than just collecting personal information. The article examines the operational workflow behind these AI-powered scams, revealing how fraudsters are using increasingly sophisticated automation to personalize their pitches and build false rapport before making financial requests.
The Original Agile Manifesto Was a Simple Set of Values
A reflection on the original Agile Manifesto argues that the software development community has overcomplicated agile methodologies with endless frameworks and theories. The author contends that the original four values and twelve principles were intentionally simple, and that the focus should be on optimizing the actual work rather than obsessing over process optimization and methodology purity.
AI in Testing: From Manual Checks to a Smart Workflow
An exploration of how AI is transforming software testing workflows, moving teams from manual test execution to intelligent, automated testing pipelines. The article covers practical applications of AI in test generation, defect prediction, and test optimization, showing how development teams can leverage AI to improve both the speed and quality of their testing processes.
How AI Actually Calls an API? Tool Calling Explained from Scratch
A technical deep dive into how AI models perform tool calling — the mechanism by which language models interact with external APIs and tools. The article explains the underlying architecture, from function calling formats to response parsing, providing developers with a clear understanding of how to implement and debug tool calling in their own AI applications.
Building Fiesta Agency: Because Apparently, “Just Make It Look Good” Wasn’t Enough
A developer shares the story of building Fiesta Agency, a creative agency platform that emerged from the frustration of clients who wanted more than just aesthetic improvements. The article details the technical challenges of building a full-service creative platform, from design tool integration to project management features, and the lessons learned about scope, client expectations, and product-market fit.
Building web-tui-kit: A Framework-Independent Browser UI with a Stable Contract
A developer describes building web-tui-kit, a browser-based UI toolkit that provides a stable interface contract independent of any specific frontend framework. The project aims to solve the fragmentation problem in web development by offering a consistent API that works across React, Vue, Svelte, and other frameworks, reducing the cost of switching between technologies.
Building vs Running: The DDL & DML Distinction Nobody Explains Simply
A clear explanation of the difference between DDL (Data Definition Language) and DML (Data Manipulation Language) in database management, concepts that are fundamental to working with databases but often poorly explained. The article uses practical examples to illustrate how DDL commands like CREATE and ALTER differ from DML commands like INSERT and UPDATE, and why understanding this distinction matters for database design and maintenance.
I Wrote a Programming Language Where You Code in Polish. Here Are 7 Things It Taught Me About Ruby.
A developer shares insights gained from creating a programming language where the syntax and keywords are in Polish, and how this exercise revealed unexpected things about Ruby’s design. The seven lessons cover topics from keyword design to language philosophy, demonstrating how building something foreign can provide fresh perspective on familiar tools.
declscope: A Linter That Brings File-Scoped Private to Go’s Flat Packages
The declscope tool addresses a long-standing complaint in the Go community by introducing file-scoped privacy to Go’s flat package structure. The linter allows developers to mark symbols as private to their containing file, providing a level of encapsulation that Go’s built-in visibility rules don’t support, while maintaining compatibility with the Go toolchain.