Musk's Young Guns Infiltrate US Government Agencies

2025-02-03
Musk's Young Guns Infiltrate US Government Agencies

WIRED reports that several young employees with ties to Elon Musk, including interns and recent graduates from companies like Neuralink, SpaceX, and xAI, are holding significant positions within US government agencies, including the Office of Personnel Management (OPM) and the General Services Administration (GSA). Their qualifications and experience have raised concerns about their competence and potential conflicts of interest. Some are even directly involved in code review and decision-making, prompting questions from within the government. This incident highlights the risks of tech elites entering government and potential vulnerabilities in agency security vetting and hiring practices.

Read more

The Ultimate Guide to File Watchers: A Comprehensive Overview

2025-02-12

This article provides a comprehensive list of file watching tools, covering various programming languages, licenses, and functionalities. From older inotify wrappers to modern tools written in Rust and Python like watchexec, the article meticulously compares each tool's pros and cons, dependencies, and Debian package support. The author also shares personal experiences and challenges encountered, such as using systemd .path units and the limitations of watchman. Developers and system administrators alike will find this resource invaluable in selecting the right file watching tool for their needs.

Read more
Development file watching

Proposal: Essential Effects for C

2025-01-20

This proposal suggests enhancing C's core type system with a type-and-effect system, requiring functions and blocks to declare the effects of their evaluation. This improves metaprogramming composability and simplifies defining MISRA-style restrictions. The system tracks and checks effects but doesn't handle dynamic effect handling. Three main effect groups are proposed: Local (e.g., local writes), Persistent (e.g., non-local writes), and Control (e.g., non-local control transfers). Static effect checking is achieved by assigning type-and-effect to each statement, expression, and function body, discarding local effects to ensure correct ordering. Best practices are suggested, like limiting the magnitude of `mut` and `vol` effects and specifying effect checks for operators and control structures. The goal is to improve C's safety, readability, and composability, especially for metaprogramming and resource management.

Read more
Development type system

ts_zip: Text Compression with Large Language Models

2024-12-30

Fabrice Bellard's new tool, ts_zip, leverages the RWKV 169M v4 language model to achieve significantly higher text compression ratios than traditional methods. While requiring a GPU and 4GB RAM, and slower than conventional compressors (up to 1 MB/s on an RTX 4090), ts_zip demonstrates impressive compression on English and other languages, including source code. Compression ratio is measured in bits per byte (bpb), with results showing superior performance to xz on datasets like enwik8 and enwik9. Deterministic model evaluation ensures consistent decompression across different hardware and software configurations.

Read more
Development text compression

Saying Goodbye to Pinboard: Migrating Links to My Own Site

2025-02-12
Saying Goodbye to Pinboard: Migrating Links to My Own Site

The author announces they're no longer updating their Pinboard bookmarks, instead migrating all links to their personal website. Reasons cited include Pinboard's lack of updates over the years, a poor mobile experience, and reservations about the Pinboard owner's views. While the author previously mirrored Pinboard links to their own site, this created a double maintenance burden. Therefore, they've decided to make their personal website the sole source of truth for links, leaving the Pinboard page as a historical record.

Read more
Misc RSS feed

Build a Tiny, YubiKey-Secured CA for Your Homelab

2025-01-19
Build a Tiny, YubiKey-Secured CA for Your Homelab

This tutorial walks you through building a secure, YubiKey-protected Certificate Authority (CA) on a Raspberry Pi. Leveraging the open-source step-ca and an optional Infinite Noise TRNG for enhanced randomness, you'll create a miniature internal ACME server for your homelab's TLS needs. The guide covers system setup, PKI creation, CA configuration, adding an ACME provisioner, and implementing systemd services for handling YubiKey removal/insertion. The result? A secure, SSH-less, tiny CA.

Read more
Development certificate authority

Apple's Closed Ecosystem Holds Back AI-Powered iOS Development

2025-02-19
Apple's Closed Ecosystem Holds Back AI-Powered iOS Development

A veteran iOS developer laments Apple's closed-source ecosystem, hindering its ability to compete with AI-powered app building platforms like lovable.dev and a0.dev. The article highlights the complexities of iOS compilation, proprietary Xcode project formats, the closed-source nature of SwiftUI, and the challenges of deploying macOS servers at scale as significant roadblocks to developing robust AI-assisted iOS development tools. In contrast, Android's open-source nature provides a significant advantage in AI app development. The author argues that Apple's long-standing neglect of developer experience has ultimately left them behind in the AI race.

Read more

Desperate Escape: A South Vietnamese Pilot's Daring Flight

2025-01-26
Desperate Escape: A South Vietnamese Pilot's Daring Flight

As South Vietnam crumbled in 1975, Air Force Major Buang-Ly, his wife, and five children, risked everything by fleeing in a small, overloaded plane. Facing enemy fire and lacking navigation, fuel, and radio, they flew towards the sea, searching for US Navy ships. Miraculously, they landed on the USS Midway, a feat made possible by the courageous decision of the carrier's captain to clear the deck despite enormous risks and potential consequences. The daring landing saved the family's lives, a testament to human resilience in the face of overwhelming odds.

Read more

Less AI Knowledge, More AI Acceptance?

2025-01-27
Less AI Knowledge, More AI Acceptance?

New research reveals a surprising finding: people with less knowledge about AI are more open to integrating it into their daily lives. This contradicts common assumptions. The study found higher AI acceptance rates in nations with lower average AI literacy. The reason? AI's ability to perform tasks previously thought exclusive to humans creates a sense of wonder and awe. Those familiar with AI's technical workings see it as a tool, not magic. Promoting AI requires balancing public understanding with maintaining enthusiasm to fully harness its potential.

Read more

Nintendo's Anti-Palworld Patent War Goes Global: US Patent Granted

2025-02-15
Nintendo's Anti-Palworld Patent War Goes Global: US Patent Granted

Nintendo secured a US patent in February 2025 for a creature-capture system, seemingly targeting Palworld. This follows a lawsuit filed in Japan against Pocketpair, the Palworld developer, for intellectual property infringement. The new patent, similar to one granted late 2024, uses subtly different wording to broaden its scope, suggesting Nintendo might expand the legal battle globally. The outcome depends on pending US patent applications, with one previously rejected but appealed by Nintendo.

Read more
Game Patent

Treewidth: A Key Parameter in Graph Theory and Its Applications

2025-01-14
Treewidth: A Key Parameter in Graph Theory and Its Applications

This article delves into treewidth, a crucial parameter in graph theory. Defined using tree decompositions, treewidth characterizes graph structure and is closely related to algorithmic complexity. The article explores multiple equivalent definitions of treewidth, its structural properties, and computational methods. It then details its broad applications in sparse numerical linear algebra, Bayesian inference, game theory, low-dimensional topology, network science, and algebraic geometry. The author also discusses advances in related width parameters and how treewidth can improve the efficiency of graph algorithms.

Read more

DeepSeek-R1: A Reasoning Model Trained via Reinforcement Learning and its Distilled Versions

2025-01-20
DeepSeek-R1: A Reasoning Model Trained via Reinforcement Learning and its Distilled Versions

DeepSeek has released its first-generation reasoning models, DeepSeek-R1. Trained via large-scale reinforcement learning without supervised fine-tuning, DeepSeek-R1 addresses issues like endless repetition and poor readability present in its predecessor, DeepSeek-R1-Zero, by incorporating cold-start data before RL. DeepSeek-R1 achieves performance comparable to OpenAI-o1 across various benchmarks. Furthermore, DeepSeek has open-sourced DeepSeek-R1 and six distilled models based on Llama and Qwen. DeepSeek-R1-Distill-Qwen-32B surpasses OpenAI-o1-mini on multiple benchmarks, setting new state-of-the-art results for distilled models. These models, along with a user-friendly API and chat interface, are available on Hugging Face.

Read more

Trump's 'Reciprocal' Tariffs: A Misunderstanding of VAT and its Impact

2025-02-15
Trump's 'Reciprocal' Tariffs: A Misunderstanding of VAT and its Impact

The Trump administration is again floating the idea of "reciprocal" tariffs on foreign countries. This plan, ostensibly a response to foreign tariffs and non-tariff barriers, fundamentally misunderstands the Value Added Tax (VAT). The article argues that equating VAT with tariffs is wrong; VAT is border-adjusted, rebating taxes on exports and imposing them on imports, resulting in a neutral economic effect. The real impediment to US competitiveness is the complex US state sales tax system, with its cascading taxes leading to "tax pyramiding" that increases costs for US businesses. Instead of raising tariffs, the article suggests reforming the US federal tax system, such as implementing full and immediate investment deductions, to boost US manufacturing competitiveness.

Read more
Tech VAT

X's Community Notes: Building a Community to Combat Misinformation

2025-01-20

The team behind X's (formerly Twitter) Community Notes shares their design process and philosophy for combating misinformation on the platform. Initially observing the difficulty in accessing accurate information online, they moved beyond traditional methods (internal review teams or media partnerships) which suffered from speed, scale, and trust issues. Inspired by Wikipedia's crowdsourced model, they developed Community Notes: users submit specific notes addressing individual posts, and an algorithm filters for notes deemed helpful across the political spectrum. This algorithm analyzes user voting history, identifying notes that bridge disagreements even among opposing viewpoints. Years of iteration and piloting led to a global rollout, significantly reducing misinformation spread and boosting user trust.

Read more
AI

Meta's AI Trained on Pirated Books: A Copyright Catastrophe

2025-01-21
Meta's AI Trained on Pirated Books: A Copyright Catastrophe

Meta is embroiled in a major copyright lawsuit after court documents revealed its AI was trained using a massive database of pirated books. Internal communications expose Meta employees admitting to using the notorious piracy site LibGen, even uploading pirated files to torrent networks. While Meta claims fair use, the blatant disregard for copyright and ethical sourcing raises serious concerns. This scandal highlights the ethical blind spots of tech giants prioritizing progress over intellectual property rights, sending shockwaves through the AI industry and beyond.

Read more
Tech

Hands-On Graphics Without X11: A NetBSD wscons Deep Dive

2025-01-17
Hands-On Graphics Without X11: A NetBSD wscons Deep Dive

This article details how to perform low-level graphics programming on NetBSD without X11 or Wayland, leveraging the wscons framework. The author demonstrates accessing and manipulating the framebuffer to draw graphics directly on the terminal and handle keyboard input. This bypasses traditional graphics systems, allowing direct hardware interaction, ideal for resource-constrained environments like embedded systems, enabling faster boot times and reduced resource consumption.

Read more
Development framebuffer

A Static Website for Professional Communication

2025-02-13
A Static Website for Professional Communication

A new static website project, "How to Professionally Say," offers alternative phrasing for common workplace communications to promote professionalism and avoid misinterpretations. Data is sourced from Instagram content creator @loewhaley. The project is in early stages, with the author seeking community feedback to refine its content and make it more broadly applicable. The project is open-source, welcoming contributions and suggestions.

Read more

Extracting an AI Model from Microsoft's Seeing AI App

2025-01-05
Extracting an AI Model from Microsoft's Seeing AI App

Security researcher Altay Akkus successfully extracted the currency recognition AI model from Microsoft's Seeing AI app. The app uses the TensorFlow Lite framework, and the model is stored encrypted within the APK file. Using the Frida framework, Altay dynamically injected code to hook the TensorFlow Lite model loading function, successfully dumping the decrypted model file. This demonstrates the vulnerability of even seemingly secure apps to AI model extraction, highlighting the importance of mobile AI model security.

Read more

Sea Turtles' Secret Navigation: It's All in the Dance

2025-02-14
Sea Turtles' Secret Navigation: It's All in the Dance

Scientists have discovered that sea turtles use Earth's magnetic field for navigation, expressing memories of food locations through a unique "dancing" behavior. Researchers trained turtles to associate specific magnetic fields with food, and the turtles responded by excitedly "dancing" when they sensed the familiar field. Published in Nature, this study reveals that turtles possess two distinct magnetoreception mechanisms: a magnetic compass and a magnetic map, suggesting these mechanisms may have evolved separately. This provides crucial insights into understanding animal magnetoreception.

Read more

Charting the Universe: Is the Cosmos Itself a Black Hole?

2024-12-24
Charting the Universe: Is the Cosmos Itself a Black Hole?

Two physicists have created a chart encompassing every known object in the universe's history, plotted by mass and size. The chart reveals that all objects reside within a triangle bounded by gravitational and Compton limits. Black holes lie on the gravitational limit, while fundamental particles are on the Compton limit. Intriguingly, the universe itself also sits on the gravitational limit, raising the question: is our universe a black hole? The chart also illustrates the universe's evolution, from the formation of fundamental particles after the Big Bang to the emergence of stars and galaxies, and points towards the exploration of unknowns like dark matter.

Read more

Bizarre Particle's Mass Depends on Travel Direction

2024-12-12
Bizarre Particle's Mass Depends on Travel Direction

Scientists have unexpectedly discovered a strange quasiparticle, a semi-Dirac fermion, in a ZrSiS material. This particle exhibits a peculiar behavior: it's massless when moving along a specific direction but gains mass when traveling in other directions. This discovery, stemming from research into the properties of quasiparticles within ZrSiS, relates to Einstein's mass-energy equivalence, E=mc². When moving at light speed in a specific direction, the quasiparticle is massless; changing direction and slowing down causes it to gain mass. The finding could potentially lead to novel applications for ZrSiS, similar to those of graphene.

Read more

Hardware-Level Network Time Security: Netnod's FPGA Implementation of NTS

2024-12-13
Hardware-Level Network Time Security: Netnod's FPGA Implementation of NTS

Following a 2019 software implementation of Network Time Security (NTS), Netnod has deployed NTS at the hardware level using FPGAs for their NTP and NTS protocols. This hardware implementation offers enhanced security, mitigating side-channel attacks and improving efficiency and scalability. While challenges existed in processing complex NTS packets, Netnod overcame them with a multi-engine parallel processing solution. Their NTS service is now in production.

Read more

Critical Apple Chip Flaws: FLOP and SLAP Attacks

2025-01-28
Critical Apple Chip Flaws: FLOP and SLAP Attacks

Researchers discovered two critical vulnerabilities, dubbed FLOP and SLAP, in Apple's M-series and A-series chips. FLOP exploits the chip's load value predictor (LVP) to steal sensitive data from Chrome and Safari browsers, including information from Gmail, iCloud, and Google Maps. SLAP, targeting primarily Safari, leverages the load address predictor (LAP) for similar data theft. Affected devices include iPhones, iPads, and Macs released since September 2021. While Apple claims to be assessing the risk, researchers have published mitigations and recommend users update their systems.

Read more

Red Hat's Minimum Viable Open Source AI

2025-02-06
Red Hat's Minimum Viable Open Source AI

Red Hat defines the minimum criteria for open-source AI as open-source-licensed model weights combined with open-source software components. This article details Red Hat's vision for open-source AI, emphasizing the importance of open licensing for model weights to facilitate community contributions and improvements. They highlight their contributions through projects like InstructLab and the Granite model family, and their commitment to building open-source AI platforms on technologies like Kubernetes and KubeFlow. Their ultimate goal is to democratize and broadly deploy open-source AI across hybrid cloud environments.

Read more

Who Owns AI-Generated Code? Legal Experts Weigh In

2024-12-24
Who Owns AI-Generated Code? Legal Experts Weigh In

The ownership of code generated by AI like ChatGPT is a complex legal grey area. Experts consulted highlight a lack of clear legal precedent, with ownership hinging on both contract and copyright law. While OpenAI disclaims ownership of generated content, in practice, ownership could fall to the user, the AI developer, or even the providers of the training data. Further complicating matters, the copyrightability of AI-generated code itself is debated; the US Copyright Office suggests the code isn't protectable, but the application incorporating it might be. The situation is legally murky and developers are urged to proceed cautiously.

Read more

Toyota's Woven City: A Real-Life Smart City Experiment

2025-01-06
Toyota's Woven City: A Real-Life Smart City Experiment

Toyota's ambitious $10 billion Woven City, a futuristic city built on the site of a former car factory in Japan, is nearing completion. Starting this summer, 100 Toyota employees will become the first residents of this “living laboratory,” testing autonomous vehicles, smart home technology, and various innovative projects. The city, designed by Bjarke Ingels, aims to eventually house 2,000 residents, powered by Toyota's hydrogen fuel cell technology. While innovative, the project faces challenges similar to Google's Sidewalk Labs, needing to balance technological advancements with resident privacy concerns. Initial residents, dubbed “Weavers,” will contribute to developing and testing new concepts, including futuristic cafe experiences, high-powered wheelchairs, and pet robots.

Read more

curl-impersonate Updated: Enhanced Browser Impersonation Capabilities

2024-12-30
curl-impersonate Updated: Enhanced Browser Impersonation Capabilities

lexiforest/curl-impersonate is an active fork of curl-impersonate, enhancing browser impersonation capabilities and supporting more versions and build targets. This project modifies curl to mimic the behavior of major browsers (Chrome, Edge, Safari, and Firefox) during TLS and HTTP handshakes, bypassing website restrictions based on fingerprinting. Updates include support for ECH, ZSTD compression, X25519Kyber768 curves, and more browser versions, offering improved command-line tools and library functions.

Read more
Development browser impersonation

EU-US Data Deal: Built on Shifting Sands?

2025-02-06
EU-US Data Deal: Built on Shifting Sands?

The EU-US Data Privacy Framework (TADPF) faces criticism for relying on potentially revocable US executive orders. The European Court of Justice previously ruled US law incompatible with EU data protection standards. Despite this, the EU Commission approved TADPF, allowing EU businesses to freely transfer data to US providers. However, a new US administration could overturn the executive orders underpinning TADPF, leaving many EU businesses in legal limbo. Experts warn EU companies should develop contingency plans to address the potential legal risks.

Read more

Keyboard Company Halts US Shipments Due to Trump Tariffs

2025-02-06
Keyboard Company Halts US Shipments Due to Trump Tariffs

Mechanical keyboard company Qwertykeys has temporarily suspended all shipments to the US due to President Trump's tariffs on Chinese goods. The 45% tariff increase, coupled with DHL's new requirement for a 50% prepayment of declared value plus a $21 processing fee per package, makes shipping unsustainable. Qwertykeys is pausing shipments for 72 hours to negotiate with DHL and other logistics providers for fairer solutions. The company also faced temporary delays in sending replacement parts due to a now-reversed US Postal Service suspension of packages from China.

Read more
Hardware trade war
1 2 510 511 512 514 516 517 518 572 573