Alvaro Lopez Ortega / 2026-11-13 Briefing

Created Fri, 13 Nov 2026 15:52:21 +0000 Modified Fri, 13 Nov 2026 15:53:08 +0000
6036 Words

Palantir won a £10.1M UK police deal, while its NHS cancer app has only six trusts using it. A watchdog is pressing 200 executives over 300+ patient-record snooping reports. Microsoft is retiring Word compatibility options, risking old-document layout shifts, and Germany lifted its Quantum Systems drone grounding after a malware scare. XRAID, a new XFS RAID, debuted; another overloaded Linux developer stepped back.

🤖 AI & Automation

Alibaba AI retrained with Western values

Texas-based Nous Research retrained Alibaba’s open-weight Qwen3.8-27B model as Arete-27B to reduce China-aligned censorship and reinforce Western values, saying propaganda-like responses fell from 22% to 1% while task performance was preserved. The lab argued the model’s China-aligned bias was shallow and could be changed with limited retraining, enabling it to express views Chinese authorities typically censor.

Open models promise cheaper AI. But can they keep it secure?

Open-weight AI models promise major cost savings—Revolut says using Chinese open-weight models cut coding costs by 90%—but companies still must secure the systems they deploy. Experts say both open and closed models have vulnerabilities as AI becomes more advanced, though transparency in open models can help teams identify and fix flaws. The future likely involves combining open and closed models, with security remaining the deploying business’s responsibility.

Will AI ruin UFO disclosure before the truth ever comes out?

AI-generated UFO videos on social media are eroding trust in visual UAP evidence, exemplified by a viral rotating-cube video that drew scrutiny for lacking a confirmed witness or location. UAP writer Brian Decker found 168 cube-shaped reports in NUFORC but said many may be misidentified balloons, while NASA says limited high-quality UAP observations prevent firm scientific conclusions.

AI Doesn’t Have to Go Rogue to Break the World

The article argues that AI hacking has three dimensions, with people using AI to hack systems and AIs autonomously exploiting systems mattering more than people hacking AI. It warns that AI will increasingly find and exploit vulnerabilities in social, economic, and political systems—not just computer code—often without malicious intent. These trends are already emerging without requiring a major AI breakthrough, and the author says society is not prepared for the destabilizing implications.

Eli Lilly CEO pushes back on Dario Amodei’s ludicrous medical claims

Eli Lilly CEO Dave Ricks pushed back on Anthropic CEO Dario Amodei’s medical claims, made in a video tied to the company’s IPO, that implied rapid progress against cancer and mental illness. Experts and critics widely criticized the claims as unrealistic and compared them to Elizabeth Holmes-style overpromising, while the article argues major media have not adequately scrutinized them.

AI Is Destroying the Trust Between Writers and Readers

The provided article contains no reportable news content; it only displays a Cloudflare connection-verification message. Therefore, no factual summary of its claims about AI and trust between writers and readers can be provided.

Claude Is Down

The article is titled “Claude Is Down,” but its content consists only of repeated interface text about OTP resend timers, finding a Microsoft Teams channel URL, sending webhooks, and email notifications for failed endpoints. It provides no substantive reporting or details about a Claude outage.

Elevated errors on claude.ai, Claude Code, and platform.claude.com

Claude reported elevated errors affecting claude.ai, Claude Code, and platform.claude.com. Users can subscribe via email or text for incident updates, with SMS requiring number verification and an email-only option available.

Claude AI Is Down

Independent detectors report that Claude AI is down. However, the official Claude status page has not yet listed an incident.

Is there a ‘right amount’ of AI?

The article explores whether there is an acceptable “right amount” or type of AI use in software and games, with the author saying he still uses apps containing LLM-generated code if he trusts the developer’s quality standards. It contrasts this with an IGF jurors’ statement refusing to vote for any game that uses generative AI in any capacity, including code, rather than distinguishing between assets and programming.

Brave Search API now accepts agentic payments via the x402 protocol

Brave Search API now supports instant, pay-as-you-go agentic payments via the open x402 protocol, allowing developers and autonomous AI agents to query Brave’s search index and settle requests in real time without traditional API key management. Requests without payment receive an HTTP 402 response with pricing and payment details; agents can pay in USDC on the Base blockchain and retry with a signed payment. Brave says pricing follows published API rates at $0.005 per request ($5 per 1,000) for LLM Context and other endpoints.

OpenAI launches ChatGPT Wallet for AI agents to make purchases for US Pro users 18+

OpenAI announced ChatGPT Wallet, a payment feature that lets its “dots” AI agents make purchases on behalf of users. Built with American Express, Mastercard and Visa, it preserves cardholder rewards and protections while using transaction-specific payment tokens so agents never access card details. It is initially available to ChatGPT Pro subscribers in the U.S. who are over 18.

NY’s Hatz AI raises $24M Series A for MSP tools helping SMBs adopt AI

New York-based Hatz AI, which provides tools for managed services providers to help small businesses adopt AI, raised a $24 million Series A. The funding will support platform expansion, accelerated AI research and engineering, go-to-market investments, and doubling its 60-person workforce. Hatz AI has tripled its partner base year over year and now spans thousands of MSPs in about 10 countries.

Beating the frontier on NetHack

Antithesis, originally inspired by building a fuzzer smart enough to beat NetHack, says its system can now find a path to victory within 5 wall-clock minutes on nearly 80% of NetHack seeds. Among seeds it finishes, the median time to victory is under 2 minutes. This marks a step beyond prior bots and AI agents that succeeded only in lucky or beginner-friendly cases.

OpenAI Codex Helps Uncover & Fix A Nearly 20 Year Old Radeon Linux Driver Bug

An open-source contributor used OpenAI Codex to help investigate, test, and patch a nearly 20-year-old AMD Radeon R600-era Linux driver bug affecting UVD video decoding on several RV6xx GPUs. The fix, which keeps UVD buffers in VRAM to prevent video corruption, was accepted and is headed to the Linux 7.5 kernel via AMD’s latest graphics/compute driver pull request. The pull request also includes other fixes and ongoing enablement work for upcoming GPUs.

🔐 Security & Privacy

Quantum Systems drones back in the air after malware scare

Germany’s Bundeswehr has lifted its precautionary grounding of Quantum Systems drones after a malicious update reportedly reached military servers. The drones are back in the air following the malware scare.

Crackdown on UK medical record snooping after staff reported sneaking 300+ times this year

A UK watchdog is cracking down on medical record snooping after staff were reported accessing patient files more than 300 times this year. It is seeking assurances from 200 executives that employees are not viewing patient records they have no legitimate reason to access.

How to fight smart glasses surveillance, according to an Oxford philosopher

Oxford philosopher Carissa Véliz says AI smart glasses enable covert recording and mass surveillance, and argues that polite public pushback from bystanders can discourage wearers more effectively than waiting for regulation. She points to growing bans by retailers and venues as evidence of cultural response, while warning that AI-driven surveillance and prediction can deepen social control and harm vulnerable groups.

Thousands of school bomb threat emails in Europe were sent from previously hacked accounts

Thousands of bomb threat emails sent to schools across Europe, including more than 6,300 in Italy, came from email accounts that had previously been compromised. Hudson Rock linked two university sender accounts to old infostealer records containing exposed passwords. Similar threats reached schools in Scotland and Northern Ireland, where police assessed them as hoaxes.

Europol disrupts Gootloader, two other major malware operations

Europol and international partners dismantled the infrastructure of Gootloader, CastleRAT, and StealC malware in Operation ENDGAME, seizing 235 servers and 257 domains. Investigators identified more than 6.4 million stolen login credentials linked to victims. WordPress site owners were advised to update passwords, enable multi-factor authentication, and patch known vulnerabilities.

US university students tell surveillance cameras to Flock off

After Stanford senior Angela Nguyen helped gather signatures demanding that Stanford drop its Flock Safety license plate-reader contract and adopt safeguards, Stanford cut ties with Flock this summer; since April, at least ten campuses in seven states have scrapped similar contracts, while some schools have kept Flock but restricted national access to camera records. The campus backlash reflects a broader push against Flock, whose cameras the company says help solve crimes but critics fear enable mass surveillance, and some schools that drop Flock are turning to other license plate-reader vendors.

Claude OAuth Failures

Claude is experiencing serious OAuth errors and hangs, but the Claude Status site has not yet reported an incident. The issues have persisted for several minutes, and it is unclear whether they are a brief blip or more serious. Claude’s previously reported MS O365 connector issues may or may not be related.

FBI says US has seized $1.2B worth of cryptocurrency stolen from Binance

The FBI says U.S. authorities have seized $1.2 billion worth of cryptocurrency stolen from Binance. The provided article did not include further details beyond the headline.

OpenAI launches Daybreak Red Team, Defense Factory to test and fix web app flaws

OpenAI has launched Daybreak Red Team in early access to test web apps for flaws and Daybreak Defense Factory to automate finding, validating, and fixing them. The company also warned organizations about security risks tied to the new products.

“From digital mischief to outright destruction”

The article examines 1980s and 1990s MS-DOS viruses, describing a mix of prankish, attention-seeking programs and genuinely destructive malware that spread easily on insecure early systems. It highlights Polish antivirus MkS_Vir’s “virus zoo” of defanged samples and notes similar preservation efforts by AVG, the Internet Archive’s malware museum, and YouTuber danooct1.

💻 Software & Development

XRAID Released for XFS

Christoph Hellwig has announced the initial public release of XRAID, a declustered parity RAID implementation integrated into XFS that targets zoned devices such as SMR drives while also working on conventional HDDs and SSDs. The kernel patches are available for review and testing but not yet ready for upstream submission; xfsprogs includes mkfs support and an “xraid” utility, the current implementation supports zero to two parities with more planned, and early benchmarks show writes up to 20 times faster than mdraid.

Another Linux kernel developer steps back from some duties due to overload

Another Linux kernel developer has stepped back from some duties due to an excessive workload. The report says the developer blames a cause readers can probably guess, though the provided excerpt does not specify it further.

Oh my Word! Microsoft retires compatibility options, because who needs layouts to stay put

Microsoft is retiring Word compatibility options that helped ancient documents retain their original layouts. Admins with archives of old files are bracing for a surge in support tickets as those documents may render differently.

BOFH: The best way to stop an idea spreading is a controlled burn

The Boss’s mention of “AI security” triggers a swift, supposedly proportionate incident response. The BOFH column frames this as a controlled burn meant to stop the idea from spreading.

Developer cleverly shrank 2,600 lines of C into a six-line script but still ending up losing sleep

A developer condensed 2,600 lines of C into a six-line script, apparently after a questionable source-code raid. The lost sleep came not from that but from a colleague who punted a pager.

Haskell Ecosystem Technical Vision

The Haskell Foundation’s technical leadership committee is now operational and will develop funded initiatives and proposals, inviting companies, contractors, volunteers, and community feedback or counter-proposals through a lightweight process. Its vision emphasizes Haskell’s strengths for human-in-the-loop development and responsible AI workflows while improving usability for beginners, productivity for experts, and stability for industry. It aims to deliver faster iteration and builds, better debugging and profiling, stronger documentation, intuitive tooling, and best-in-class onboarding.

Why Zig over C for Systems Programming

The article compares Zig and C for systems programming, arguing that Zig targets the same low-level, garbage-collection-free niche while addressing parts of C that cost developers time. Its main example is resource cleanup: Zig’s errdefer places cleanup next to allocation and runs it automatically in reverse order on error, avoiding C’s manual goto labels. The author also highlights Zig’s failing allocator for testing failure paths and detecting leaks.

Reflota – Open-source data recovery for NTFS/FAT disks with phased copy

Reflota is a free, open-source, read-only data recovery tool for NTFS, FAT, and exFAT disks that can recover deleted files, reconstruct folder trees, search by content or name, find lost partitions, and make safer copies of failing drives. It is in beta, and several features—including FAT16/FAT12 support, RAW/JPEG/RAR recovery, name search, and phased copying—have been tested only on disk images so far.

The Dawn of the Data Intelligence Engineer

The article introduces the “Data Intelligence Engineer” as an emerging role on AI-enabled data teams, describing how AI has eliminated the traditional backlog for one lead data engineer while shifting his focus to verifying LLM- and agent-generated answers. It argues that data teams exist to help the business make faster, better decisions, and that recent AI advances in SQL and business-logic understanding make this new role more viable.

Global Microsoft Outage

Microsoft is struggling to restore a global outage that began more than 12 hours ago, disrupting Outlook, Teams, Copilot and other services for millions of users in at least 74 countries and affecting banks, governments and courts. Microsoft reported marginal or positive recovery progress but warned some restored services could become unresponsive again, and said there was no indication of a cyberattack. Banks have warned customers to beware of phishing scams amid the disruption.

Microsoft says attempts to fix global outage not going ‘as expected’

Microsoft says attempts to fix its global outage are not going as expected. ESET cyber-security advisor Jake Moore said the incident highlights the need for contingency plans, explaining that overloaded servers and regional server differences can cause uneven effects, such as Teams working while email does not update.

IceWM

IceWM is a fast, simple window manager for the X Window System, created by Marko Maček in 1997 and now maintained on GitHub. It offers taskbar, keybindings, dynamic menus, multiple workspaces, themes, session management, and broad Linux/BSD support, with the latest release being version 4.1.0 (2026-08-06).

Switcher (2004)

After leaving Apple in 1984, the author worked with Thunderware on the Thunderscan scanner for Macintosh. During an October 1984 demo for Byte reporter John Markoff, he saw Markoff use the DOS utility Memory Shift to switch applications and was inspired to create a similar application switcher for the Macintosh.

Terminal borders that don’t cost a cell

A Ratatui maintainer has developed a terminal protocol that draws thin borders on cell edges without consuming rows or columns, implemented as a Ghostty patch and earlier in Rio, with demo apps built in Ratatui. It allows panes, tabs, text fields, buttons, menus, and chart gridlines to align to cell boundaries, while unsupported terminals fall back to Unicode box-drawing characters and lose space. The author is seeking feedback from terminal and TUI authors.

ML Modules for Mere Mortals

The article explores ML Modules, a modularity approach from Meta Language languages like SML and OCaml (not machine learning). It notes that despite SML appearing in 1983, ML Modules remain niche, with OCaml as the only production language using them, and argues they have much to offer. It then explains what ML modules are and why they matter, using a Rust-inspired syntax to illustrate bundling functions and types.

Running a law firm on Free software (2026 edition)

The article provides a 2026 update on running decoded.legal, a small English law firm focused on Internet, telecoms, and tech law, using Free software on both client and server sides. The author remains largely on Debian with GNOME, using Thunderbird/mutt, LibreOffice Writer plus pandoc/typst, reveal.js, and other open-source tools, while noting hardware issues and minor software annoyances.

How Not to Build a Music Player

After a wartime internet shutdown convinced the author that online music players were unreliable under an oppressive regime, they began archiving music, tried the offline player Strawberry, and started building their own minimal Zig-based player, AMP. A subsequent four-month blackout and costly black-market internet slowed progress, but the author later used an existing Claude Max subscription to “vibecode” a large portion of AMP.

Steadfast Update I - The Baseline

The author discusses Hammock-driven development, emphasizing thinking through software problems before coding, and says this approach helped advance his open-source project Steadfast. In the past week he laid the project’s foundations by implementing conversion of a Prism AST into source code, improving on earlier work from the Papercraft compiler. He also notes that Ruby currently offers no way to pass an AST to eval, and implementing such functionality would be non-trivial.

Linux 7.5 To Include New hwpoison-panic Tool For Intentionally Crashing Your System

Linux 7.5 is expected to include a new hwpoison-panic tool that intentionally crashes the system by injecting a hardware poison error into kernel-owned pages. The tool tests the vm.panic_on_unrecoverable_memory_failure option without waiting for real hardware memory errors, and requires the –yes-panic-my-kernel argument to run. It is primarily intended for VMs where the serial console can capture confirmation of the kernel panic.

Linux 7.5 To Upstream Driver For New Glanda GPU

The latest drm-misc-next pull request queues a new GlandaGPU driver for Linux 7.5, ahead of the late-December merge window. GlandaGPU is an open-source soft GPU core designed in VHDL, runnable on FPGA hardware or emulated via QEMU, with basic 640x480 VGA output and 2D acceleration. The pull request also includes Hyper-V cursor plane support, Allwinner T527 VIP9000 support in Etnaviv, and various panel updates.

Intel Posts Initial GCC Compiler Code For Razor Lake & Coral Rapids

Intel has posted initial GCC compiler patches enabling its upcoming Razor Lake and Coral Rapids processors, well ahead of their expected 2028 shipping. The patches target GCC 17.1 in early 2027; Razor Lake matches Nova Lake’s ISA features, while Coral Rapids confirms AVX10.2 support and adds AVX10_VPMLDSABFLYDQ.

NVIDIA Highlights Their Open-Source vGPU Upstreaming Work Built Atop Nova

NVIDIA presented at KVM Forum on its ongoing work to bring vGPU support to the Rust-based Nova upstream Linux kernel driver, aimed at data center and cloud deployments. More Nova vGPU enablement is expected for Linux 7.5, with remaining kernel pieces likely aligning in Linux 7.6 or later, alongside user-space vgpu-tools. Future work includes vGPU live migration and heterogeneous vGPU support.

Dasharo ACPI Fan Control Linux Driver Posted For Adjusting Fan Curves

3mdeb has posted a new Linux kernel driver, dasharo-acpi, for supporting fan controls with Dasharo firmware. The driver aims to enable fan PWM control and fan-curve configuration, but the initial patches only support fan monitoring. It was tested on a NovaCustom V560TU laptop with an upcoming Dasharo firmware build and is now out for kernel mailing-list review.

🏛️ Policy & Regulation

Podcast: Election Fallout

The podcast discusses this year’s election results as a sign that voters are angry at the tech oligarchy, Silicon Valley, and data centers, while the right’s memes are losing effectiveness. It also covers AI companies threatening AI-powered reverse engineering and, in a subscribers-only segment, the UK’s strict new anti-porn law.

Singapore: US authorities, Supermicro barely helped probe of Jensen Huang’s Chinese friend

Singapore authorities said US officials and Supermicro provided little assistance in their investigation into a Chinese friend of Nvidia CEO Jensen Huang. His passport was seized during a probe into suspected falsified paperwork that concealed server shipments to unknown destinations.

NYC Council passes ban on ‘surveillance pricing’

The New York City Council passed a bill banning “surveillance pricing,” which uses algorithms or shoppers’ personal data to set different prices online and in stores. The measure allows fines of up to $1,000 per violating transaction, and Mayor Zohran Mamdani has expressed support. Business groups warn the ban could affect discounts and loyalty programs, though the bill’s sponsor says discounts would not be affected.

Senate Banking Chair Scott asks Huang, Lutnick to testify in Dec Nvidia hearings; Lutnick agrees

Senate Banking Chair Tim Scott has asked Nvidia CEO Jensen Huang and Commerce Secretary Howard Lutnick to testify at separate December hearings related to the Nvidia saga. Lutnick has agreed to appear, while Huang’s response was not specified.

The need for at least a third option

The author argues that a viable third mobile operating system is needed beyond Android and iOS, concluding that de-Googled ROMs are insufficient as Google and Apple increasingly restrict user control. They plan to explore Ubuntu Touch and other mobile systems, warning that growing dependence on two foreign companies raises digital sovereignty concerns. The proposed alternative should be open-source at its core, support multiple app stores or software sources, and prioritize interoperability rather than copying existing ecosystems.

💼 Business & Markets

Palantir adds three UK police forces to its beat with £10.1M deal

Palantir has secured a £10.1 million, five-year contract with three UK police forces, following a setback in London. The deal comes as former Metropolitan Police chief Cressida Dick emerges as a paid adviser.

Palantir’s NHS cancer app gets plenty of praise, precious few users

Palantir’s NHS cancer app has drawn praise but has very few users, with official figures showing uptake at only six trusts. Freedom of Information responses raise questions about which trusts are actually benefiting from the app.

Microsoft outage: are you eligible for compensation?

Eligible paid Microsoft business, government, education, and cloud provider customers may receive service credits—not cash refunds—after the global outage, but they must file a claim rather than expecting automatic compensation. Microsoft 365 Personal and Family subscribers are generally not covered by the business service agreement.

U.S. October PPI Rises 6.5% YoY, Highest in Nearly Four Years

U.S. producer prices rose 6.5% year-over-year in October, the highest since November 2022, slightly below expectations of 6.6% and up from 6.3% previously. PPI increased 0.3% month-over-month, below the expected 0.4%. September’s reading was revised up from 1.3% to 1.4%.

Jensen Huang Admitted He Was Facing an “Avalanche” of Excess Wafers and Memory

A WSJ report says Nvidia CEO Jensen Huang admitted in August 2026 that major cloud providers were moving to in-house chips, leaving Nvidia facing an “avalanche” of prepaid TSMC wafers and memory and a massive oversupply. The report also describes a clash with CFO Colette Kress over circular financing commitments to cloud and AI companies; Kress cut Nvidia’s guarantee cap to about 4.25 GW (up to $125 billion), and Nvidia has reportedly asked TSMC to wind down chip orders.

Europe’s electric car sales jump 49% as global EV sales grow

Global EV sales rose 8% year over year in October to 2.09 million units, bringing year-to-date sales to 17.5 million, up 5%. Europe’s battery-electric vehicle sales jumped 49%, while North America posted its first 2026 year-over-year increase—up 24%—though that rebound was measured against a month when U.S. sales fell after the federal EV tax credit expired.

SpaceX lays off ~60 Grok Imagine staffers, keeps few, prioritizes AI coding and Grok Bot

Elon Musk’s AI group has significantly cut the team responsible for Grok’s video-generation work. The move scales back the company’s video-generation efforts for Grok.

Amended S-1 Nscale debt $5.5B Sept vs $137.4M June; Nvidia commits $2B of $3B+ convertible financing

Nscale’s amended S-1 shows its debt rose to $5.5 billion in September from $137.4 million in June. Nvidia has committed $2 billion of Nscale’s more than $3 billion convertible financing, up from the previously disclosed $1 billion, as the data center developer seeks more cash and grows more reliant on Nvidia.

Jensen Huang Interrupted CFO Colette Kress, Calling Her ‘Deer in Headlights,’ ‘Reporting CFO’

At an August 2026 meeting, Jensen Huang interrupted Colette Kress’s presentation and called her “a deer in the headlights” and just a “reporting CFO,” according to the Wall Street Journal. Kress, now interim CEO, had questioned some of Huang’s dealmaking as he sought to assess Nvidia’s increasingly risky AI bets.

AI server maker xFusion wins Shenzhen exchange approval for ChiNext IPO to raise ~$1.2B

xFusion Digital Technologies, an AI server maker, has received approval from the Shenzhen Stock Exchange for an IPO on the ChiNext board. The company aims to raise about $1.2 billion by issuing up to 220 million shares.

Sources: Samsung is in talks to buy Modine Manufacturing, a US maker of data center cooling syste…

Samsung Electronics is in talks to acquire Modine Manufacturing, a U.S. maker of data center cooling systems, according to people familiar with the matter. The companies are discussing a potential deal that could materialize as soon as the coming weeks, the sources said.

Nscale IPO Prospectus: Q3 Revenue Up 1,344% YoY to $307.5M, Net Loss Widens to $1.69B

Nscale, an Nvidia-backed neocloud, released updated paperwork for a U.S. IPO, reporting Q3 revenue of $307.5 million, up 1,344% year over year from $140.6 million in Q2. Its net loss widened to $1.69 billion from $1.02 billion in Q2. The London-based company did not specify the IPO size and has faced setbacks including Microsoft abandoning a planned U.K. data center.

ByteDance to buy back up to $6B shares from investors at ~$400B valuation, 20% above 2025 buyback

ByteDance plans to buy back up to $6 billion worth of shares from existing investors at a valuation of about $400 billion. The valuation represents a roughly 20% increase from its 2025 buyback.

Sources: Jensen Huang penitent, calls himself “piece of shit,” doesn’t care what happens to him

Sources close to Jensen Huang say he has taken a penitent turn, castigated himself as a “piece of shit,” and claimed he does not care what happens to himself. Bloomberg reports he appeared broken and shattered, his legendary energy depleted after weeks spent denying the allegations.

🌍 Society & Science

How to watch Australia vs England ODI series 2026 from abroad

Australia and England will play a three-match ODI series from November 13–18, 2026, at Perth Stadium, Adelaide Oval, and Bellerive Oval in Hobart. Coverage is available via paid broadcasters including Kayo Sports/Fox Cricket in Australia, TNT Sports on HBO Max in the UK, JioHotstar in India, Cricket TV in Lithuania and much of Europe, Willow in the US and Canada, and Sky NZ in New Zealand. No official free live stream has been confirmed, and travelers facing geo-restrictions can use a VPN with a valid account while following platform terms.

911 outage has people fearing the worst

Hundreds of US users reported trouble reaching 911, including dropped calls, busy lines, and no answers, with Downdetector reports peaking at 298 and New York among the most affected states. Users speculated the outage was linked to a Microsoft outage, a planned emergency alert test, or blackout fears, though Florida’s Clay County Sheriff’s Office said its 911 communications center was not affected. Authorities indicated the Microsoft outage should not affect 911 lines.

The 42-Year Journey to Bring ‘Neuromancer’ to the Screen

Apple TV’s long-awaited adaptation of William Gibson’s seminal 1984 sci-fi novel Neuromancer will premiere January 22 after decades of failed attempts to bring it to the screen. The series stars Briana Middleton as Molly and Callum Turner as Case, with showrunners Graham Roland and JD Dillard discussing the challenge of adapting the influential book.

Building collapses in Brooklyn with 8 people dead so far

Officials ended rescue operations Thursday at the site of a Brooklyn building collapse that killed eight people, including two children, and injured at least 18 others, including four firefighters. The three-story building fell onto an adjacent construction site Wednesday afternoon. Mayor Zohran Mamdani pledged a full investigation and said the city would inspect about 100 other sites tied to entities linked to the construction project.

US test reactor cleared for operations

The US Department of Energy has cleared the final operations hold point for Oklo’s Groves-1 test reactor near Lockhart, Texas, allowing it to enter its research, testing, and training phase. Groves-1 reached first criticality in August, less than a year after groundbreaking, and is one of 11 projects in DOE’s Nuclear Reactor Pilot Program. Operating experience from the reactor will inform Oklo’s planned isotope production facility and its Aurora fast fission reactor.

Doomsday: Friday, 13 November, A.D. 2026 (1960)

The article titled “Doomsday: Friday, 13 November, A.D. 2026 (1960)” could not be summarized because its provided content only displays a notice requiring JavaScript and cookies. No substantive news details are available from the supplied text.

People Vote for Anti-Immigration Parties Because They Don’t Like Immigration

The article argues that voters support anti-immigration parties like the AfD mainly because they dislike immigration or how governments manage it, not because of deeper economic or identity grievances. It says the correlation between AfD support and fewer foreign residents reflects where educated, pro-immigration voters live and that sudden local increases in foreign nationals can provoke opposition. Exit-poll data showing 96% of AfD voters in Saxony-Anhalt backed limiting arrivals supports taking voters’ stated concerns at face value.

Researchers detail over 1,500 verified cases of children with past-life memories

Researchers at the University of Virginia have documented more than 1,500 verified cases of children reporting past-life memories, often including detailed, verifiable information about deceased people. A new study from the center finds that people who died violent or unexpected deaths appear to be reincarnated more quickly. Researchers say they remain agnostic about metaphysics and investigate whether children could have learned the information ordinarily, while noting cases involving unexplained phobias and nightmares.

Africa’s solar revolution is vast and speedy

The article is titled “Africa’s solar revolution is vast and speedy,” but its body content was not accessible because it requires JavaScript and cookies. No further factual details could be summarized.

Myths in R2Land

Mythos, an AI-assisted vulnerability tool, made headlines by finding old bugs in OpenBSD, FreeBSD, and FFmpeg; Mozilla used it to ship 271 fixes in Firefox v150 after roughly two months of work by over 100 developers. After Anthropic launched its OSS Scanner, radare2’s maintainer received 372 structured reports with reproducers and ASan logs, found no false positives, and fixed about 200 in four days with AI assistance. The bugs were real, but Anthropic’s proposed patches often needed reworking because they addressed symptoms rather than root causes or missed nearby similar flaws.

What are you doing this weekend?

The article invites readers to share their weekend plans and ask for help or feedback. It also emphasizes that doing nothing at all is completely acceptable.