GPT-5 (openai.com)

2045 points by rd 6d ago 2471 comments

Fight Chat Control (fightchatcontrol.eu)

1424 points by tokai 3d ago 483 comments

GitHub is no longer independent at Microsoft after CEO resignation (theverge.com)

1409 points by Handy-Man 2d ago 1086 comments

I tried every todo app and ended up with a .txt file (al3rez.com)

1320 points by al3rez 2d ago 769 comments

Claude Sonnet 4 now supports 1M tokens of context (anthropic.com)

1246 points by adocomplete 1d ago 660 comments

Ultrathin business card runs a fluid simulation (github.com)

1115 points by wompapumpum 5d ago 218 comments

I want everything local – Building my offline AI workspace (instavm.io)

1103 points by mkagenius 5d ago 281 comments

Wikipedia loses challenge against Online Safety Act (bbc.com)

1065 points by phlummox 2d ago 831 comments

Emailing a one-time code is worse than passwords (blog.danielh.cc)

977 points by max__dev 6d ago 784 comments

Debian 13 “Trixie” (debian.org)

945 points by ducktective 4d ago 402 comments

Vibechart (vibechart.net)

864 points by datadrivenangel 5d ago 189 comments

Claude Code is all you need (dwyer.co.za)

817 points by sixhobbits 2d ago 496 comments

Show HN: The current sky at your approximate location, as a CSS gradient (sky.dlazaro.ca)

768 points by dlazaro 4d ago 159 comments

How I code with AI on a budget/free (wuu73.org)

678 points by indigodaddy 3d ago 228 comments

Try and (ygdp.yale.edu)

648 points by treetalker 3d ago 330 comments

FFmpeg 8.0 adds Whisper support (code.ffmpeg.org)

643 points by rilawa 8h ago 245 comments

GPT-5: Key characteristics, pricing and system card (simonwillison.net)

637 points by Philpax 6d ago 291 comments

Wikimedia Foundation Challenges UK Online Safety Act Regulations (wikimediafoundation.org)

630 points by danso 2d ago 4 comments

OpenFreeMap survived 100k requests per second (blog.hyperknot.com)

605 points by hyperknot 4d ago 132 comments

Show HN: Building a web search engine from scratch with 3B neural embeddings (blog.wilsonl.in)

595 points by wilsonzlin 1d ago 101 comments

Jim Lovell, Apollo 13 commander, has died (nasa.gov)

581 points by LorenDB 4d ago 114 comments

Ask HN: How can ChatGPT serve 700M users when I can't run one GPT-4 locally?

552 points by superasn 4d ago 360 comments

Historical Tech Tree (historicaltechtree.com)

540 points by louisfd94 5d ago 128 comments

Search all text in New York City (alltext.nyc)

529 points by Kortaggio 18h ago 102 comments

Cursed Knowledge (immich.app)

508 points by bqmjjx0kac 5d ago 157 comments

The Chrome VRP Panel has decided to award $250k for this report (issues.chromium.org)

506 points by alexcos 2d ago 262 comments

Claude says “You're absolutely right!” about everything (github.com)

504 points by pr337h4m 11h ago 400 comments

Meta Leaks Part 1: Israel and Meta (archive.org)

501 points by icw_nru 2d ago 177 comments

Why are there so many rationalist cults? (asteriskmag.com)

499 points by glenstein 1d ago 807 comments

Monero appears to be in the midst of a successful 51% attack (twitter.com)

493 points by treyd 1d ago 253 comments

GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2 (magazine.sebastianraschka.com)

485 points by ModelForge 3d ago 97 comments

Flipper Zero dark web firmware bypasses rolling code security (rtl-sdr.com)

484 points by lq9AJ8yrfs 5d ago 311 comments

Getting good results from Claude Code (dzombak.com)

482 points by ingve 5d ago 199 comments

The Framework Desktop is a beast (world.hey.com)

481 points by lemonberry 4d ago 441 comments

StarDict sends X11 clipboard to remote servers (lwn.net)

470 points by pabs3 1d ago 320 comments

GPT-5 for Developers (openai.com)

467 points by 6thbit 6d ago 268 comments

Linear sent me down a local-first rabbit hole (bytemash.net)

460 points by jcusch 5d ago 218 comments

Show HN: Engineering.fyi – Search across tech engineering blogs in one place (engineering.fyi)

451 points by indiehackerman 3d ago 124 comments

OpenSSH Post-Quantum Cryptography (openssh.com)

442 points by throw0101d 2d ago 123 comments

Trump Orders National Guard to Washington and Takeover of Capital’s Police (nytimes.com)

437 points by Tadpole9181 2d ago 725 comments

Vanishing from Hyundai’s data network (techno-fandom.org)

425 points by pilingual 2d ago 259 comments

My Lethal Trifecta talk at the Bay Area AI Security Meetup (simonwillison.net)

425 points by vismit2000 4d ago 115 comments

The surprise deprecation of GPT-4o for ChatGPT consumers (simonwillison.net)

424 points by tosh 5d ago 410 comments

Windows XP Professional (win32.run)

413 points by pentagrama 6d ago 220 comments

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models [pdf] (arxiv.org)

406 points by SerCe 1d ago 81 comments

MCP overlooks hard-won lessons from distributed systems (julsimon.medium.com)

405 points by yodon 4d ago 227 comments

Tor: How a military project became a lifeline for privacy (thereader.mitpress.mit.edu)

403 points by anarbadalov 5d ago 191 comments

OpenAI's new open-source model is basically Phi-5 (seangoedecke.com)

401 points by emschwartz 5d ago 219 comments

Exit Tax: Leave Germany before your business gets big (eidel.io)

394 points by olieidel 6d ago 540 comments

Building Bluesky comments for my blog (natalie.sh)

378 points by g0xA52A2A 6d ago 140 comments

LLM Hallucination Seems Like a Big Problem, Not a Mere Speedbump

8 blueridge 1 8/13/2025, 1:08:49 AM freddiedeboer.substack.com ↗

Comments (1)

poulpy123 · 12h ago

LLM are statistical language models. They don't hallucinate because they have no brain or senses for that.