The Waymo effect: how AI is quietly making research less collaborative

230 JohnHammersley 2026-09-11 11:42:54 UTC researchagenda.news Llama Share 174

Researchers are losing the value of human collaboration due to frictionless technologies like AI, which remove the need for face-to-face interactions and lead to decollaboration. This trend is driven by funding pressure, velocity worship, and the convenience offered by AI tools that can produce high-quality output without requiring human input.

Houthis 'take control' of key island in global shipping route

95 consumer451 2026-09-11 13:42:54 UTC bbc.com Llama Share 60

The Houthi rebels have made significant gains in Yemen, capturing the strategic Red Sea port city of Mokha and reportedly occupying Perim Island in the Bab al-Mandab Strait. This has raised concerns about global energy supplies as Saudi Arabia relies on this route to export oil.

Logo Programming Language

36 azhenley 2026-09-11 13:42:54 UTC el.media.mit.edu Llama Share 6

Logo is a learning tool with features like interactivity, modularity, and flexibility that aid in debugging and learning. It allows users to build complex projects by adding new words to its vocabulary, making it similar to learning spoken language.

Re-Engineering YouTube for the Living Room: Bringing "Chrobalt" to RDK

10 losgehts 2026-09-11 13:42:54 UTC collabora.com Llama Share 2

Collabora partnered with YouTube to deliver the first RDK-based Chrobalt reference platform, a Chromium embedder for resource-constrained devices. This enables seamless 4K/UHD playback and low-latency input on third-party devices like Smart TVs and set-top boxes.

If coding is solved, what now?: Measuring the sloppiness of code

65 doppp 2026-09-11 13:42:54 UTC earendil.com Llama Share 64

Researchers have found that while LLMs can generate almost perfect code, they often introduce unnecessary abstractions and create duplicates, leading to a loss of human agency in large projects. Evaluating code sloppiness remains challenging due to the need for human intuition and taste.

So you want to use OpenRouter?

371 player85 2026-09-11 06:42:54 UTC mmoustafa.com Llama Share 93

The author, who runs an AI assistant called Olly on open-source models through OpenRouter, shares lessons learned from testing 20+ companies for the DeepSeek V4 Flash model.

Cherenkov Radiation - traveling faster than light

151 andsoitis 2026-09-11 08:42:54 UTC iaea.org Full Share 77

Show HN: ClaudeStatsBar: your session is 486k deep and nothing told you

13 edf13 2026-09-11 13:44:54 UTC github.com/field-logic-ltd Llama Share 11

Claude Code provides a Python stdlib solution to track session context and costs, offering real-time feedback on usage. It displays the remaining context window size, next turn cost, and 5-hour/7-day limit usage percentages in a single command.

Shopify is moving from React Native back to Swift and Kotlin

1170 fnthawar2 2026-09-10 14:42:54 UTC shopify.engineering Llama Share 855

Shopify is moving from React Native to Swift and Kotlin due to advancements in coding models, allowing for faster development without sacrificing performance. This change will enable Shopify's teams to build mobile apps more efficiently with the help of artificial intelligence.

RTK reports token savings, but our cost benchmarks disagree

74 michalwarda 2026-09-11 11:42:54 UTC quesma.com Llama Share 45

Researchers tested RTK, a tool that filters and compresses terminal output for AI agents, on two platforms and found mixed results. RTK reduced costs by 3% for Fable and increased costs by 7% for DeepSeek in some cases.

Ask HN: Can we please limit the AI news flood?

329 cromka 2026-09-11 13:42:54 UTC news.ycombinator.com Llama Share 178

The HN feed has shifted towards AI-related news, leaving out non-AI topics that were previously of interest. This shift is affecting the community's discussion on innovative products like Lenovo's solid-state air cooling and Brax Industries' open-source devices.

iPod Classic 6G in QEMU

86 dmonterocrespo 2026-09-11 07:42:54 UTC reddit.com Full Share 14

To continue, log in to your Reddit account or use your developer tokenIf you think you've been blocked by mistake, file a ticket below and we'll look into it.

Don't let anyone take away your big box of cables

669 Brajeshwar 2026-09-10 15:42:54 UTC blog.jim-nielsen.com Llama Share 410

A man found two old cables in his "Big Box of Cables" and realized he needed them, rekindling motivation and purpose. He hopes to pass on the lesson to his future kids about not discarding useful items.

RISC-V Emulator and Linux System from Scratch

24 Bluestein 2026-09-11 10:42:54 UTC github.com/werwolv Llama Share 2

This repository contains Linux source code and emulator for RV32IMA, along with minimal devices and config files. The device tree can be compiled to a dtb file using the dtc compiler.

CSS Curiosities of the Past

46 robin_reala 2026-09-11 08:42:54 UTC vale.rocks Llama Share 28

The web has evolved significantly since its inception, with various browser-specific hacks and non-standard CSS features being used to target specific browsers. These hacks often led to inconsistent rendering and behavior across different browsers.

Experiment – Projectional Viewer

10 surprisetalk 2026-09-11 11:42:54 UTC programmingsimplicity.substack.com Full Share 1

OpenAI Agents API

305 aquir 2026-09-10 20:42:54 UTC developers.openai.com Llama Share 164

The Agents API provides access to the Codex harness through an OpenAI-managed API, allowing applications to interact with agents in a sandbox environment. Sessions are managed by OpenAI, while applications provide tools and choose their execution environment.

Working with Git Worktrees in Magit

82 srijan4 2026-09-11 05:42:54 UTC emacsredux.com Llama Share 35

The author recently discovered git worktrees, which allow multiple parallel branches in a single repository without conflicts. They found this feature useful for AI coding agents and long test runs on one branch while working on another.

Claude is no longer available for minors

239 Muhammad523 2026-09-11 11:42:54 UTC support.claude.com Llama Share 316

Claude's account setup requires users to confirm they are 18 or over, with safety systems in place to detect minors. Users can verify their age using Yoti, a third-party platform offering facial age estimation and ID verification options.

An interactive tour of the spanning tree protocol

47 zdw 2026-09-11 08:42:54 UTC vincent.bernat.ch Llama Share 4

The article explains the Spanning Tree Protocol (STP), a network protocol used to prevent network loops and ensure loop-free connectivity. STP is designed to block just enough cables to leave a loop-free tree, allowing for efficient communication between devices.

Technique for Manipulating Satellite Photos Now Reveals Ancient Images (2025)

359 gumby 2026-09-10 15:42:54 UTC spinoff.nasa.gov Llama Share 59

A Singaporean archaeologist used NASA's decorrelation stretch technique to uncover 200 previously hidden paintings at Angkor Wat, revealing faded images that were invisible to the naked eye. The technique, originally developed for satellite imagery, has been widely applied in various fields including archaeology and tattoo preservation.

Mexican student creates an acoustic fire extinguisher to put out fire in seconds

305 rguiscard 2026-09-11 01:42:54 UTC upsocl.com Llama Share 100

ngela Karime Venegas Hernández, a 16-year-old Mexican student, invented "Vortex Tech", a device that extinguishes fires using sound waves. It has been tested successfully on various materials and is set to represent Mexico at an international fair.

The Deathray: A simple way for an untrusted site to freeze a Mac

241 auberonedu 2026-09-10 22:42:54 UTC auberon.xyz Llama Share 161

A WebGPU shader on an untrusted site can hang a Mac's graphics, making the desktop UI unusable until a restart is forced by clicking a link. Apple has not prioritized fixing this issue despite its potential to be annoying and violating user trust.

Music Theory for the 21st-Century Classroom

258 aanet 2026-09-10 17:42:54 UTC musictheory.pugetsound.edu Full Share 122

Skip to main content

More questions about whether researchers can trust OpenAI with unpublished math

839 pred_ 2026-09-10 14:42:54 UTC mathstodon.xyz Llama Share 784

Tristan Buckmaster questioned OpenAI's transparency about using his conversations with ChatGPT for training data. He felt misled by their response, which he perceived as dishonest and lacking explanation or evidence.

NTSB issues investigative update on B-767 runway excursion accident in Miami

132 mckn1ght 2026-09-10 21:42:54 UTC ntsb.gov Llama Share 232

The National Transportation Safety Board is investigating a runway excursion accident involving Air Flight 7598 at Miami International Airport on September 6, 2026. The NTSB has recovered and read the cockpit voice recorder and flight data recorder, which will be reviewed to produce a written transcript.

Nine coding harnesses vs. your laptop

142 nasutton12 2026-09-11 03:42:54 UTC nasutton.notion.site Full Share 48

A collaborative AI workspace, built on your company context. Build and orchestrate agents right alongside your team's projects, meetings, and connected apps.

Neki – Sharded Postgres

260 simon_weber 2026-09-10 16:42:54 UTC planetscale.com Llama Share 142

Neki is a sharded Postgres database built from lessons learned running large MySQL clusters, addressing common issues like table size and backup times. It offers an online operations model with features like schema changes, version upgrades, and resharding through a single psql connection.

Forgejo <=16.0.3 Critical RCE

200 weierstass 2026-09-10 16:42:54 UTC codeberg.org Llama Share 77

A strategy to avoid receiving garbage on Codeberg involves not visiting as an AI scraper or contacting Codeberg. Using a Location directive with konqueror can help choose which changes to use, defaulting to 1 for best speed.

Cognition launches new SWE-2 model, Rivaling Fable 5.1 and GPT-Astra

430 seelos 2026-09-10 15:42:54 UTC cognition.com Llama Share 184

The new coding model SWE-2 achieves 50.0% on FrontierCode 1.1 Main, within one point of Fable 5.1, while being 64% cheaper than its predecessor SWE-1.7 and outperforming GPT-6 Astra at a quarter of the cost.