Sublinear-Space Zero-Knowledge Proofs: A Breakthrough

2025-09-24
Sublinear-Space Zero-Knowledge Proofs: A Breakthrough

Researchers have developed SSZKP, a sublinear-space zero-knowledge proof system with a Rust implementation using KZG/BN254. It achieves O(√T) memory usage for a trace of length T, employing blocked IFFTs and streaming accumulators. SSZKP uses standard KZG commitments without buffering entire polynomials, offering a highly efficient approach for building scalable zk systems. The project includes complete prover and verifier CLI tools and end-to-end test scripts.

Read more
Development

LLM 0.26: Large Language Models Get Terminal Tooling

2025-05-27
LLM 0.26: Large Language Models Get Terminal Tooling

LLM 0.26 is out, bringing the biggest feature since the project started: tool support. The LLM CLI and Python library now let you give LLMs from OpenAI, Anthropic, Gemini, and local Ollama models access to any tool representable as a Python function. The article details installing and using tool plugins, running tools via the command line or Python API, and shows examples with OpenAI, Anthropic, Gemini, and even the tiny Qwen-3 model. Beyond built-in tools, custom plugins like simpleeval (for math), quickjs (for JavaScript), and sqlite (for database queries) are showcased. This tool support addresses LLM weaknesses like mathematical calculations, dramatically expanding capabilities and opening up possibilities for powerful AI applications.

Read more
Development Tool Support Plugins

UK Adults with ADHD Experience 6-9 Year Life Expectancy Reduction

2025-01-24

A matched cohort study using UK primary care data reveals a significant life expectancy deficit for adults diagnosed with ADHD. Analyzing data from over 9.5 million individuals across 792 general practices between 2000 and 2019, researchers found a reduction of 6.78 years for men and 8.64 years for women compared to the general population. This shortened lifespan is likely attributed to modifiable risk factors and unmet support needs for both ADHD and co-occurring mental and physical health conditions. The findings highlight a critical unmet need for improved support and treatment for adults with ADHD.

Read more

Torque: A Lightweight Meta-Assembler for Any Architecture

2025-04-16

Torque is a lightweight meta-assembler providing the tools to write programs for any processor architecture. Addressing shortcomings of existing embedded processor assemblers—poor documentation, clunky languages, bloat, single OS support—Torque uses integers, bit sequences, labels, and powerful macros to emulate any assembly language. Processor instruction encoding is defined via macros, allowing programming for any processor using only Torque and the datasheet. Source code is available at code.benbridle.com/torque-asm.

Read more
Development meta-assembler

Microsoft Office's Startup Boost: Faster Launch, Slower PC?

2025-05-01
Microsoft Office's Startup Boost: Faster Launch, Slower PC?

Microsoft is introducing a new "Startup Boost" feature for Office, pre-loading apps like Word and Excel when Windows starts to speed up their launch times. However, this could slow down overall computer performance. The feature will only be enabled on PCs with at least 8GB of RAM and 5GB of free disk space. While users can disable it in Word's settings or Task Scheduler, the move raises questions about whether Microsoft should prioritize improving Office's efficiency instead of relying on pre-loading. The update will initially roll out to Word in mid-May and later to other Office applications.

Read more

Parisian Panic: A Cascade Delete Disaster in Production

2025-06-21
Parisian Panic: A Cascade Delete Disaster in Production

A software engineer working at Joe AI, a Paris-based real estate startup, accidentally deleted a user record in the production database, triggering a cascade delete that wiped out three months of crucial data. While some data was recovered by upgrading to a paid Supabase plan, the incident highlighted the risks of directly manipulating production databases and the critical need for robust backup strategies. The disaster ultimately spurred the team to improve their development workflow, setting up local Supabase instances, leading to increased efficiency. The experience underscores the importance of learning from mistakes and embracing a culture of risk-taking and iterative improvement.

Read more

Meteor Burst Communication: A Resurgent Tech for National Security?

2025-04-20

This paper explores the potential of meteor burst communication (MBC) for national security. For decades, researchers have investigated using the ionized trails left by meteors in the atmosphere for communication. Early systems like JANET and COMET were deployed, and the US established SNOTEL and AMBCS. Advances in microprocessors have significantly improved MBC's data rates and reliability, presenting a renewed potential for national security applications such as command and control, intelligence, surveillance, and reconnaissance. While MBC has limitations like intermittent service and noise susceptibility, its flexibility and inherent anti-jamming capabilities make it a valuable supplementary communication method.

Read more

Darklang Goes Open Source: A Journey from Closed Source to Community Embrace

2025-06-16
Darklang Goes Open Source: A Journey from Closed Source to Community Embrace

Darklang, a programming language initially launched as a closed-source, hosted-only platform, has announced that it's open-sourcing all its repositories. This shift stems from product maturity, user feedback, and a change in technical direction. Initially, Darklang's closed-source approach aimed to ensure sustainability and deliver unique features like safe code migration and unified deployment. However, with product maturity, user requests for more openness, and the emergence of local-first development and new business models (e.g., paid team collaboration and AI tools), Darklang chose to embrace open source to foster community growth and the platform's long-term evolution. While some technical challenges remain, such as licensing implications in package management, the open-sourced Darklang will be more transparent and accessible.

Read more
Development

Ugly Gerry: A Font Fighting Gerrymandering

2025-05-30
Ugly Gerry: A Font Fighting Gerrymandering

Ugly Gerry is a typeface whose letters are shaped like US congressional districts, a protest against gerrymandering. Created by Ben Doessel and James Lee for RepresentUs, the font's intentionally grotesque design aims to highlight the unfairness of manipulated district lines. While dubbed "the world's most revolting font," its provocative design earned it a 2020 ADC Award for typography, successfully bringing attention to a crucial political issue.

Read more

Cracking Mendel's Last Mystery: Genes for Three Pea Traits Finally Identified

2025-04-25
Cracking Mendel's Last Mystery: Genes for Three Pea Traits Finally Identified

Gregor Mendel's groundbreaking pea plant experiments, conducted over 160 years ago, laid the foundation for genetics. However, the genes responsible for three of his seven observed traits remained elusive. A new study published in Nature uses modern genomic sequencing and sophisticated computational analysis to finally identify these genes, revealing how they control pod color, pod shape, and flower branching. This breakthrough not only solves a long-standing puzzle but also paves the way for advancements in pea genomics and plant breeding.

Read more
Tech peas

Servo vs. Ladybird: A Battle of New Browser Engines

2025-03-26
Servo vs. Ladybird: A Battle of New Browser Engines

This article compares Servo and Ladybird, two projects aiming to revolutionize the browser engine landscape. Servo, initially backed by Mozilla, transitioned to the Linux Foundation due to funding issues and is now developed by Igalia with an undisclosed but significant funding source. Ladybird, started by Andreas Kling, relies on Patreon, GitHub sponsorships, and ad revenue, and has grown into an independent project with 7 full-time engineers, boasting substantial donations. In web standards compliance tests, Ladybird slightly edges out Servo, although Servo excels in CSS tests. Performance-wise, Servo significantly outperforms Ladybird, but both lag behind mainstream browsers. Both are open-source, but target different audiences and development models; Servo emphasizes embeddability, while Ladybird focuses on the browser itself.

Read more

The 50-Year-Old Mystery of Internet Packet Size

2025-04-18

This article delves into the enduring question of optimal internet packet size. From RFC 791's initial suggestion of 576 octets to today's default of 20-1500 octets, packet size has been a key trade-off in network design. Tracing the evolution of Ethernet, it explains the relationship between minimum packet size and collision detection, and the balance between maximum packet size and transmission efficiency. Jumbo frames and Path MTU discovery are also discussed, concluding that 46-1500 octets remains a reasonable range for the public internet, a choice that has persisted for nearly 50 years.

Read more

AI in OS: Hype Over Substance?

2025-04-12
AI in OS: Hype Over Substance?

Microsoft, Apple, and Google are aggressively pushing AI integration into their operating systems, such as Microsoft's Copilot and Apple Intelligence. However, the article argues this is more hype than practical benefit. Users prefer stable, private, and customizable OSes without unnecessary bloat, ads, or invasive AI features. While AI assistants have value in specific niches (like programming), forcing their integration into the OS sacrifices user experience and facilitates greater data collection by tech companies. The ideal OS is stable, private, lightweight, and customizable, with AI tools offered as optional standalone apps, not core OS functions.

Read more

Falklands Starlink Shutdown: A Tech-Political Standoff

2025-02-08
Falklands Starlink Shutdown: A Tech-Political Standoff

Hundreds of Falkland Islanders face internet outages as their Starlink 60-day roaming period expires. Despite a successful petition garnering widespread support and a parliamentary committee endorsing Starlink's use, its operation remains illegal due to Sure's telecommunications monopoly. While the government issued a limited number of VSAT licenses, Starlink lacks formal approval, leaving many users in a legal gray area. The article urges the government to declare a national emergency to swiftly resolve the issue, prevent further outages, and highlights the government's swift response to similar emergencies during the COVID-19 crisis.

Read more

From Arc to Dia: The Browser Company's AI Pivot

2025-05-27
From Arc to Dia: The Browser Company's AI Pivot

The Browser Company announces its pivot from its browser project, Arc, to a new AI-powered browser, Dia. The article explains that despite Arc's loyal user base, it fell short of expectations due to growth limitations, hesitation in embracing AI, and communication issues. The author admits Dia isn't a simple iteration of Arc but rather a response to the future of browsers—AI will fundamentally reshape how we interact with computers, and traditional browsers will be replaced by AI browsers. Dia aims to address Arc's shortcomings, offering a simpler, faster, and more secure experience. The article also touches on Arc's future, stating it won't be open-sourced immediately but will continue to be maintained.

Read more
Development Pivot

Cuttlefish Communicate Using Elaborate Arm Movements and Water Vibrations

2025-05-07
Cuttlefish Communicate Using Elaborate Arm Movements and Water Vibrations

Scientists have discovered that cuttlefish use distinct arm movements to communicate, employing a multi-sensory system involving both visual cues and water vibrations. Researchers identified four distinct arm gestures, combined with skin color changes, that appear to convey different meanings. Remarkably, cuttlefish seem to perceive these signals differently depending on their orientation, similar to how humans perceive faces. Furthermore, experiments revealed that cuttlefish can detect these signals through water movement, likely utilizing their lateral line and statocysts. This dual-channel communication system adds a new dimension to our understanding of cephalopod intelligence and animal communication.

Read more

Open-Source Tool LVTShift: Model Your City's Land Value Tax

2025-06-05
Open-Source Tool LVTShift: Model Your City's Land Value Tax

This blog post details using the open-source tool LVTShift to model the impacts of a land value tax (LVT). The author showcases analyses of South Bend and Syracuse, demonstrating how LVTShift simulates various LVT policies (revenue-neutral, different tax burden shifts, etc.) and their effects on city residents and economies. The post thoroughly explains data acquisition, processing, model building, and analysis, including code examples and data sources. Readers are encouraged to model their city's LVT using LVTShift and share their results.

Read more
Development

Nigerian Village Lights Up Thanks to Solar Mini-Grid

2025-03-15
Nigerian Village Lights Up Thanks to Solar Mini-Grid

Two remote Nigerian villages, Mbiabet Esieyere and Mbiabet Udouba, previously reliant on kerosene lamps and expensive generators, now enjoy reliable, affordable electricity thanks to a solar mini-grid installed by Prado Power in 2022. The project, initially met with skepticism, has transformed lives and boosted local businesses. A barbershop owner's monthly electricity costs dropped dramatically, and a cassava farmer's weekly income increased fivefold. This success story highlights the potential of mini-grids to address Africa's energy access challenge and underscores the importance of supportive policies, community engagement, and external funding in driving renewable energy adoption.

Read more
Tech mini-grid

22 Sailors Rescued After EV Fire Engulfs Cargo Ship

2025-06-06
22 Sailors Rescued After EV Fire Engulfs Cargo Ship

A fire broke out on the Morning Midas, a cargo ship carrying 3,000 vehicles (800 of which were electric), approximately 304 miles south of Adak, Alaska. The fire, believed to have originated from electric vehicle batteries, proved difficult to extinguish due to the salt water's conductive properties causing short circuits and thermal runaway in the batteries. All 22 crew members were rescued, but the ship is expected to continue burning until specialized firefighting crews arrive. This incident highlights the significant challenges posed by electric vehicle battery fires at sea.

Read more

LLMs: The Trash Disguised as the Future

2025-07-04

This article reflects on PayPal's early monopoly on online payments and Bitcoin's initial promise of decentralized payments, ultimately becoming a speculative tool. The author connects this to the current proliferation of LLMs (Large Language Models), arguing that LLMs haven't delivered promised convenience and innovation, but instead generate useless and misleading information, lowering the bar for creation and devaluing creative work. The author criticizes LLMs for their low-quality, unreliable output and their negative impact across industries, leading to a flood of 'content' and a decline in quality. The author concludes by advocating for genuine creation and sharing of meaningful work.

Read more
(eev.ee)

AWS S3: How Cheap Hard Drives Power a Massively Scalable Storage System

2025-09-24
AWS S3: How Cheap Hard Drives Power a Massively Scalable Storage System

This article unveils the astounding scale and underlying technology of Amazon S3. S3 leverages inexpensive HDDs, overcoming the limitations of slow random I/O through massive parallelization, erasure coding, and clever load balancing techniques (like the 'power of two choices'). This enables millions of requests per second, ultra-high throughput, and exceptional availability. S3's data storage strategy incorporates random data placement, continuous rebalancing, and the smoothing effect of scale to avoid hot spots. Parallelization at the user, client, and server levels further boosts performance. Ultimately, S3 has evolved from a backup and image storage service to a foundational component of big data analytics and machine learning infrastructures.

Read more
Tech

YouTube Cracks Down on AI-Generated Fake Movie Trailers

2025-05-16
YouTube Cracks Down on AI-Generated Fake Movie Trailers

YouTube is taking action against channels creating fake movie trailers using AI-generated content. Channels like Screen Trailers and Royal Trailer, which amassed millions of views with misleading trailers splicing real clips and AI-generated material, have had their ad revenue suspended. Hollywood studios are reportedly pushing YouTube to redirect this revenue. This crackdown highlights the challenges of regulating AI-generated content, protecting intellectual property, and combating misinformation on online platforms. The action follows an investigation revealing the deceptive nature of these trailers and their significant viewership.

Read more
Tech

Vertical Sharding: A Nightmare?

2025-04-13
Vertical Sharding: A Nightmare?

The author recounts their experience with vertical sharding (functional sharding), highlighting its pitfalls. While it alleviates database load, it fragments the application, forcing the application layer to handle joins and queries that should be handled by the database. This significantly increases code complexity and maintenance overhead, and reduces system availability. Using humor and an uptime formula, the author shows how vertical sharding lowers system stability, ultimately delaying product roadmaps and demoralizing engineers. The article concludes by introducing PgDog, an open-source project aiming to solve Postgres sharding.

Read more
Development vertical sharding

Saying 'Please' and 'Thank You' to ChatGPT Costs OpenAI Millions

2025-04-20
Saying 'Please' and 'Thank You' to ChatGPT Costs OpenAI Millions

OpenAI CEO Sam Altman revealed that user politeness, specifically saying "please" and "thank you" to ChatGPT, costs the company tens of millions of dollars in electricity. While Altman claims it's money well spent, the revelation highlights the massive energy consumption of AI. A survey shows 70% of users are polite to AI, partly fearing a robot uprising. However, the debate rages on: does politeness improve responses, and is it worth the environmental cost? Some argue polite prompts yield better, less biased results, improving AI reliability.

Read more
AI

Solving Sudoku in tmux: A Madcap Python Compiler Project

2025-02-11
Solving Sudoku in tmux: A Madcap Python Compiler Project

Following up on his previous project compiling Python to run on tmux, the author has now created a Sudoku solver entirely within tmux. Eschewing arrays and strings, he cleverly leverages tmux's variables and keybindings, mapping each Sudoku cell to a tmux option. The solver brute-forces its way through all possibilities, resulting in extremely low efficiency. However, the project showcases the surprising capabilities of tmux and the author's ingenuity, a testament to the hacker spirit.

Read more
Development sudoku

Nuclear-Powered Pacemakers: A Forgotten Chapter in Medical History

2025-05-31

Have you ever heard of nuclear-powered pacemakers? In the past, some pacemakers utilized plutonium-238 as a power source, generating electricity via thermoelectric effects to stimulate the heart. These devices were remarkably durable, able to withstand gunshots and even cremation. Despite emitting low radiation doses, between 50 and 100 people in the US were still using them around 2003. Upon a patient's death, the pacemakers were retrieved to recover the plutonium. This article showcases a Medtronic nuclear pacemaker with its plutonium removed, measuring approximately 2.75 inches in diameter and donated by the Los Alamos National Laboratory.

Read more
Hardware

arXivLabs: Community Collaboration on New arXiv Features

2025-06-01
arXivLabs: Community Collaboration on New arXiv Features

arXivLabs is a framework for collaborators to develop and share new features directly on the arXiv website. Individuals and organizations involved share arXiv's values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners who adhere to them. Have an idea to improve the arXiv community? Learn more about arXivLabs.

Read more
Development

XChat's End-to-End Encryption: The Juicebox Security Flaw

2025-06-09
XChat's End-to-End Encryption: The Juicebox Security Flaw

Matthew Garrett exposes security vulnerabilities in X's (formerly Twitter) new end-to-end encrypted messaging protocol, XChat. XChat uses the Juicebox protocol to store user private keys, distributing them across three servers. However, these servers are all controlled by X, meaning X can access all user keys, undermining end-to-end encryption. The article delves into Juicebox's mechanics and potential risks, highlighting critical flaws in XChat's deployment. User private keys are vulnerable to arbitrary access by X, leading to the recommendation to avoid using XChat.

Read more
Tech

AI Winter Bites: NZ Tech Job Market Stagnant, Immigrants Face Headwinds

2025-06-03
AI Winter Bites: NZ Tech Job Market Stagnant, Immigrants Face Headwinds

New Zealand's tech sector is facing a downturn in 2025, with layoffs at major companies and the expansion of AI exacerbating job pressures. Microsoft cut 6,000 jobs, and New Zealand's health sector also slashed IT roles. A 12-year veteran software engineer from China, James Zhang, struggled to find work in New Zealand, citing ageism in the Chinese tech industry and visa challenges. While job ads have slightly increased, competition remains fierce, with immigrants facing additional hurdles. Many are forced to upskill or pursue further education.

Read more
Tech

Model Context Protocol (MCP): A New Standard for Building Powerful LLM Applications

2025-04-13

This article introduces the Model Context Protocol (MCP), an open protocol for building enterprise-grade Large Language Model (LLM) applications. MCP solves the problem of a lack of standardization in integrating LLMs with enterprise tools, allowing frameworks like LangChain to seamlessly integrate with various data sources and tools such as databases and GitHub. The article details MCP's core components (MCP server, client, and host), installation setup, and Python hands-on demonstrations. These include building a LangChain application to calculate simple and compound interest using Ollama, and interacting with multiple MCP servers using both stdio and sse transport modes. With MCP, LLM applications can more effectively leverage enterprise data and tools for more powerful functionality.

Read more
1 2 25 26 27 29 31 32 33 596 597