Tech News

The SpaceX IPO made history. Is the excitement still there?

A month on from the firm's stock market debut, the reality of how SpaceX currently makes money has seemed to come into…

Ask HN: Add flag for AI-generated articles

A Hacker News user proposes adding a flag to mark AI-generated articles as an indicator rather than a ranking penalty,…

Gameboy is back ( no signup)

10 years of Pokémon Go and the millions still trying to catch 'em all

Michael Steranka, vice president at the mobile game's owner Scopely, says it has always been about bringing people…

Billion Dollar PDFs

A "Billion Dollar PDF" is a document that crystallized a compelling narrative at a pivotal moment and redirected…

Circular Obstacle Pathfinding (2017)

The A* pathfinding algorithm, commonly used for grid-based navigation, can work on any graph structure to find optimal…

Designing and assembling my first PCB

The author describes their first experience designing and assembling a custom PCB, starting from learning Arduino basics…

Architecture Description Languages [pdf]

I cannot summarize this article because the provided text is only the raw PDF file code and metadata, not the actual…

Why Vanilla JavaScript

Cyberpunk Comics, Manga and Graphic Novels

This article provides a curated index of cyberpunk comics, manga, and graphic novels, listing seminal works spanning…

A Speed Limit for Computers

The article draws a parallel between Ivan Illich's 1973 theory that excessive energy consumption reduces equity and…

Show HN: Codebase Posters – turn any Git repo into generative poster art

Codebase Posters is a tool that transforms any Git repository's commit history into generative poster art by running a…

I trained a 113M-parameter earthquake LLM from absolute scratch

A developer trained a 113-million-parameter language model specialized in earthquake science from scratch, using six…

I made a free Islamic prayer app to quit saying "I'll pray in 5 minutes"

A developer created "Better Prayer," a free Islamic prayer app designed to combat procrastination by sending reminders…

Kode Dot Programmable pocket device for makers, pentesters and geeks

Kode Dot is a programmable handheld device featuring an ESP32 dual microcontroller, AMOLED touchscreen, and built-in…

Ask HN: What Are You Working On? (July 2026)

This is a Hacker News discussion thread where developers share their current projects (July 2026).

Since Chronium 148, Math.tanh is now fingerprintable to link underlying OS

Starting in Chrome 148, the Math.tanh() function has become a reliable fingerprinting technique to identify the user's…

Show HN: Adaptive Recall, persistent memory for AI assistants over MCP

Adaptive Recall is a memory API for AI assistants that goes beyond standard vector search by using cognitive science and…

Show HN: Zen Mode – a global focus mode for macOS

Zen Mode is a macOS tool that creates a global focus mode across all applications using a single hotkey (Option + Cmd +…

The State of MCP Security [pdf]

I cannot provide a summary of this document because the PDF content is corrupted or improperly encoded, displaying only…

Flash-MSA: Accelerating Million-Token Training with Sparse Attention Kernels

Researchers have developed Flash-MSA, the first open-source training kernels for MiniMax Sparse Attention that enable…

Tiny Emulators

Tiny Emulators is a collection of web-based emulators for classic computers and gaming systems from the 1980s, including…

Irish datacenters now guzzle 23% of the country's electricity

Irish datacenters consumed 23% of the country's electricity in 2025, up from 20% in 2023, with consumption increasing by…

Defining new Jax types with hijax

Hijax is an experimental JAX feature that allows developers to define custom types beyond standard arrays and pytrees,…

Neocities: Create your own free website

Neocities is a free platform that lets users create and host personal websites using an in-browser HTML editor or their…

The One-Step Trap (In AI Research)

Rich Sutton identifies the "one-step trap" as a critical flaw in AI research: the mistaken assumption that learning only…

I love LLMs, I hate hype

The author is enthusiastic about AI progress and LLMs but criticizes the surrounding hype—particularly doomsday…

Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k

A comparative analysis of Claude Code and OpenCode found that Claude Code consumes significantly more tokens before…

Can We Understand How Large Language Models Reason?