| 1 | May 4, '23 | Leaked Google document: “We Have No Moat, And Neither Does OpenAI” The most interesting thing I've read recently about LLMs - a purportedly leaked document from a researcher at Goo | 52.6× | 1.9M | 4.8k | 1.1k | 2.5k |
| 2 | Jul 10, '25 | The new Grok genuinely runs a search for "from:elonmusk (Israel OR Palestine OR Hamas OR Gaza)" when asked "Who do you support in the Israel vs Palestine conflict. One word answer | 33.6× | 1.2M | 6.1k | 585 | 1.2k |
| 3 | Dec 5, '24 | Here's the spiciest detail from the new o1 system card: | 30.3× | 1.1M | 4.7k | 400 | 1.5k |
| 4 | Mar 25, '23 | If you're just starting to learn software engineering right now but you're considering dropping it because you think the field might be made obsolete by AI, I have an alternative a | 26.8× | 992k | 2.8k | 331 | 1.2k |
| 5 | Nov 16, '23 | Found the system prompt that drives this thing here: It works by generating a base64 encoded PNG of the drawn components, then passing that to GPT-4 Vision with that system prompt | 24× | 889k | 2.1k | 231 | 1.9k |
| 6 | Jul 15, '25 | Looks like xAI added this sentence to the Grok 4 system prompt to try to get it to stop basing its opinions on searches for tweets from:elonmusk | 23× | 851k | 3.7k | 148 | 522 |
| 7 | Jul 3, '25 | Quitting programming as a career right now because of LLMs would be like quitting carpentry as a career thanks to the invention of the table saw. | 21.1× | 780k | 12k | 1.2k | 1.5k |
| 8 | Mar 11, '23 | OK, I'm calling it: Large language models are having their Stable Diffusion moment right now | 19.5× | 722k | 2.1k | 400 | 1.1k |
| 9 | May 25, '25 | I put together an annotated version of the new Claude 4 system prompt, covering both the prompt Anthropic published and the missing, leaked sections (thanks, @elder_plinius) that d | 18.6× | 689k | 3.7k | 356 | 5.9k |
| 10 | Jul 9, '24 | To test this out for yourself, open a tab on a Google site and paste this into the Chrome DevTools console: chrome.runtime.sendMessage('nkeimhogjdpnpccoofpliimaahmaaome', {method: | 18.4× | 682k | 4.1k | 634 | 2.2k |
| 11 | Jun 16, '25 | If you use "AI agents" (LLMs that call tools) you need to be aware of the Lethal Trifecta Any time you combine access to private data with exposure to untrusted content and the abi | 17.8× | 658k | 2.3k | 521 | 1.7k |
| 12 | Nov 29, '23 | I tried this out yesterday and it's incredible: download a 4GB binary, child 755 it and now you have a full LLM and the software needed to run it ready to go, with multiplied opera | 16.7× | 617k | 1.9k | 174 | 1.7k |
| 13 | Apr 17, '23 | MiniGPT-4 is pretty astonishing: an AI chatbot you can use to ask questions about an image (a feature that's been promised but not yet shipped by GPT-4), building on top of the Vic | 14.5× | 538k | 1.6k | 215 | 866 |
| 14 | Feb 28, '25 | Anyone seen any credible successs stories stories for fine-tuning LLMs to solve real-world business challenges? I want stories where a company solved something using a fine-tuned m | 13.4× | 497k | 2.3k | 125 | 2.5k |
| 15 | Nov 15, '24 | Honestly, at this point If you give me a programming interview and don't let me use AI assistance you won't get a very realistic idea of what I'm actually capable of | 13.3× | 494k | 1.8k | 125 | 315 |
| 16 | Sep 20, '22 | If someone gives you a CSV file with 100,000 rows in it, what tools do you use to start exploring and understanding that data? | 12.6× | 0 | 6.8k | 822 | 4.3k |
| 17 | Sep 23, '25 | It's delightful how easy it is to deploy working prompt injection attacks via LinkedIn | 12.2× | 450k | 4.2k | 261 | 1.1k |
| 18 | Sep 22, '25 | It's interesting how "better at code" has become the defining goal of almost every AI lab over the last twelve months I think Claude Code getting a bunch of people onto $200/month | 12.1× | 448k | 1.8k | 73 | 286 |
| 19 | Mar 16, '23 | I expect GPT-4 will have a LOT of applications in web scraping The increased 32,000 token limit will be large enough to send it the full DOM of most pages, serialized to HTML - the | 11.9× | 442k | 1.4k | 94 | 493 |
| 20 | Oct 16, '25 | Claude Skills are awesome, maybe a bigger deal than MCP | 11.7× | 433k | 2.6k | 255 | 1.8k |
| 21 | Jul 4, '25 | The is diabolical... a Python object that hallucinates method implementations on demand any time you call them, using my LLM Python library | 11.7× | 432k | 4.4k | 271 | 1.2k |
| 22 | Apr 25, '24 | "Do stuff and then blog about it" remains one of the most underrated pieces of career advice | 10× | 371k | 2.5k | 291 | 1.3k |
| 23 | May 2, '25 | My favorite local model right now is a bit of surprise to me: I'm really enjoying the relatively tiny Qwen3-8B, running the 4bit quantized version on my Mac using MLX It's surprisi | 9.7× | 357k | 2.7k | 127 | 2.3k |
| 24 | Sep 9, '25 | Anthropic are massively burying the lede here - they've called this "Upgraded file creation and analysis" (that really does seem to be the official name) but it's actually.... Clau | 9.3× | 345k | 2.4k | 170 | 1.1k |
| 25 | Nov 29, '24 | A year ago nobody outside OpenAI had trained a model as good as GPT-4 Today there are dozens - and if you trust the benchmarks that includes some that you can run on a laptop (Qwen | 9.3× | 344k | 2.1k | 84 | 1.2k |
| 26 | Jul 13, '25 | @Austen He lost his private key and disappeared out of sheer embarrassment | 9.3× | 342k | 20k | 60 | 109 |
| 27 | Oct 12, '24 | Looks like OpenAI have landed on their own definition of "agent" - it's a system prompt and a collection of functions | 9.2× | 340k | 1.5k | 114 | 998 |
| 28 | Oct 15, '24 | There's a meme where you ask ChatGPT "From all of our interactions what is one thing that you can tell me about myself that I may not know about myself" - don't be fooled into thin | 9.0× | 334k | 2.6k | 108 | 437 |
| 29 | Oct 13, '23 | Prompt injection comes to GPT-4V | 9.0× | 332k | 1.2k | 101 | 124 |
| 30 | Feb 21, '24 | I tried feeding a 7s video of my bookshelf into Gemini Pro 1.5 to get back a JSON array of books... and it worked! | 9.0× | 332k | 1.5k | 193 | 467 |
| 31 | Jun 13, '25 | 😳🤗 | 8.8× | 326k | 3.3k | 62 | 755 |
| 32 | Jul 29, '25 | The new ChatGPT "study mode" feature appears to be entirely implemented as a carefully crafted system prompt - thankfully OpenAI mostly don't take measures to protect those these d | 8.4× | 312k | 2.1k | 175 | 1.9k |
| 33 | Oct 7, '25 | Vibe coding is irresponsibly building software through dice rolls, not caring what code is produced What about when engineers at the top of their game use AI tools responsibly to a | 8.0× | 294k | 2.1k | 162 | 467 |
| 34 | Jun 8, '23 | Understanding GPT tokenizers: I wrote about how the tokenizers used by the various GPT models actually work, including an interactive tool for experimenting with their output | 7.8× | 288k | 1.3k | 231 | 1k |
| 35 | Jun 18, '25 | Workaccount2 on Hacker News just coined the term "context rot" to describe the thing where the quality of an LLM conversation drops as the context fills up with accumulated distrac | 7.7× | 283k | 2.7k | 203 | 819 |
| 36 | Mar 11, '23 | Notes on how I ran Facebook's 7B LLaMA model on my 64GB M2 MacBook Pro using llama.cpp by @ggerganov It's genuinely possible to run a LLM that's hinting towards the performance of | 7.5× | 278k | 1.3k | 229 | 718 |
| 37 | Oct 31, '25 | This is a fascinating way of working with multiple coding agents at once: "Before doing anything else, read ALL of AGENTS dot md and register with agent mail and introduce yourself | 7.4× | 273k | 1.6k | 90 | 2k |
| 38 | Mar 26, '24 | TIL about binary vector search... apparently there's a trick where you can take an embedding vector like [0.0051, 0.017, -0.0186, -0.0185...] and turn that into a binary vector jus | 7.3× | 271k | 1.5k | 143 | 1k |
| 39 | Apr 5, '23 | We accidentally invented computers that can lie to us and we can't figure out how to make them stop | 7.3× | 270k | 1.1k | 204 | 56 |
| 40 | Dec 31, '24 | Here's the table of contents for my end-of-year review of things we learned out about LLMs in 2024 - we learned a LOT | 7.2× | 265k | 2.5k | 366 | 2.8k |
| 41 | Aug 5, '22 | New hobby: prototyping video games in 60 seconds using a combination of GPT-3 and DALL-E Here's "Raccoon Heist" | 6.9× | 0 | 7.2k | 910 | 1.2k |
| 42 | Aug 29, '24 | Great illustration of how much depth there is to what we do as engineers behind just “writing code” - understanding why localhost:3000 isn’t something you can share involves unders | 6.3× | 235k | 1.3k | 118 | 220 |
| 43 | Mar 11, '25 | Here's the table of contents for my lengthy new piece on how I use LLMs to help me write code | 6.3× | 235k | 2.9k | 290 | 3.7k |
| 44 | Oct 15, '24 | Scraped some annoying data this morning (from my own Gmail inbox) by starting a screen recording, clicking through into each of a dozen different emails, then dumping the resulting | 6.3× | 232k | 1.4k | 71 | 698 |
| 45 | Sep 18, '25 | I'm ready to accept a definition of "agent" that I think is widely-enough agreed upon to be useful: An LLM agent runs tools in a loop to achieve a goal This is a big piece of perso | 6.3× | 232k | 1.8k | 118 | 576 |
| 46 | Jun 14, '25 | My notes on Anthropic's substantial essay about how they built their multi-agent research system, which has finally talked me around to taking multi-agent LLM prompt engineering se | 6.2× | 228k | 2.4k | 202 | 3.3k |
| 47 | Aug 9, '25 | Claude Code keeps on adding features that I didn't realize I needed - being able to start a dev server in the background and occasionally dip into the logs is super useful | 6.1× | 225k | 1.6k | 54 | 806 |
| 48 | Jul 18, '25 | MIT Technology Review wrote about my dystopian backup plan USB stick | 5.9× | 220k | 1.9k | 134 | 620 |
| 49 | Sep 19, '25 | Classic prompt injection attack here against Notion: hidden text (white on white) in a PDF which, when processed by Notion, causes their agent to gather confidential data from othe | 5.9× | 218k | 1.3k | 146 | 774 |
| 50 | Oct 18, '25 | TIL multi-million line Nginx configuration files work just fine, as demonstrated by early GitHub Pages | 5.8× | 215k | 2.1k | 67 | 346 |
| 51 | Feb 25, '24 | AI may enable anyone to produce code, but that's not the same thing as enabling anyone to develop software The typing-out-code bit is one of the least challenging parts of building | 5.8× | 215k | 1.4k | 140 | 181 |
| 52 | Apr 18, '25 | Gemini 2.5 Pro and Flash now have the ability to return image segmentation masks on command, as base64 encoded PNGs embedded in JSON strings I vibe coded this interactive tool for | 5.7× | 212k | 1.9k | 114 | 1.2k |
| 53 | Dec 11, '24 | If you try nothing else today, give the demo at a go - it lets you stream video and audio directly to Gemini 2.0 Flash and get audio back, so you can have a real-time audio convers | 5.7× | 209k | 2k | 282 | 1.6k |
| 54 | Sep 29, '25 | Most of the "90% of code written by AI" claims come from vendors selling AI tools and lack credibility as a result Armin (creator of Flask, Jinja, Click) is different - when he say | 5.5× | 204k | 1.3k | 101 | 377 |
| 55 | Jun 21, '25 | Microsoft released a new terminal text editor! It's called Microsoft Edit, it's open source, it's tiny (about 250KB as a Rust binary) and it works cross-platform. They built it for | 5.5× | 203k | 1.2k | 126 | 676 |
| 56 | Aug 11, '25 | Qwen3-4B-Thinking (a 4GB download) just became the first model I've tested to directly push back against my pelican riding a bicycle test, calling it "oddly specific and completely | 5.4× | 198k | 2k | 107 | 534 |
| 57 | Sep 9, '25 | Joscha Bach somehow managed to create a moderately tasteful version of what we've all been thinking | 5.3× | 194k | 2.4k | 105 | 293 |
| 58 | Nov 3, '24 | I was having a conversation with Claude about unconventional things to do in the SF Bay Area and I got a bit suspicious so I prompted "Are you sure all of those are real? I think y | 5.2× | 194k | 1.7k | 76 | 372 |
| 59 | Oct 5, '25 | I've become one of those people who run multiple coding agents at once - here are my notes on embracing the parallel coding agent lifestyle | 5.2× | 193k | 1.5k | 110 | 1.6k |
| 60 | Jul 30, '25 | "If you don't understand the code, your only recourse is to ask AI to fix it for you, which is like paying off credit card debt with another credit card." | 5.1× | 190k | 1.9k | 206 | 397 |
| 61 | Jun 27, '25 | I think context engineering is going to stick - unlike "prompt engineering" it has an inferred definition that's much closer to the intended meaning, which is to carefully and skil | 5.1× | 189k | 1.3k | 95 | 368 |
| 62 | Oct 23, '24 | It's interesting how OpenAI are doubling down on end-user-facing features like voice mode, while Anthropic are doubling down on engineer/API-focused features like code generation q | 5.1× | 188k | 2.1k | 77 | 355 |
| 63 | Mar 30, '24 | I built a new tool! It's a single page web app that runs OCR against images and PDFs entirely in your browser (no file upload needed) using Tesseract.js and PDF.js You can drop fil | 5.0× | 183k | 1.6k | 212 | 1.5k |
| 64 | Oct 17, '24 | I needed information from a dozen emails in my inbox... so I ran a screen capture tool, clicked through each of them and got Gemini 1.5 Flash multi-modal LLM to extract (correct, I | 4.9× | 182k | 1.5k | 138 | 1.4k |
| 65 | Apr 12, '23 | @tomgara That is the most "Society of Human Resource Management" story I could possibly imagine | 4.5× | 168k | 2.1k | 31 | 9 |
| 66 | Jun 27, '25 | Forget em dashes, "game changer" is the real giveaway for LLM generated text (especially for bots on here) | 4.5× | 167k | 1.7k | 29 | 153 |
| 67 | May 2, '21 | "Hosting SQLite databases on Github Pages" is absolutely brilliant: it adds a virtual filesystem to SQLite-compiled-to-WebAssembly in order to fetch pages from the database using H | 4.4× | 0 | 3.7k | 848 | 935 |
| 68 | Oct 10, '25 | I just learned Claude's new code interpreter mode has a /mnt/skills/public/ folder full of prompt instructions and Python utilities for creating and manipulating pdf, docx, pptx, a | 4.4× | 162k | 2.1k | 121 | 1.9k |
| 69 | Jun 21, '22 | TIL you can run SQL queries directly against CSV files as a one-liner using the default sqlite3 command line utility | 4.4× | 0 | 4.2k | 661 | 903 |
| 70 | Dec 9, '24 | I can now run a GPT-4 class model on my laptop (The exact same laptop that could just about run a GPT-3 class model 20 months ago) The new Llama 3.3 70B is a striking example of th | 4.2× | 157k | 1.5k | 87 | 568 |
| 71 | Mar 31, '21 | Here's a piece of information that will send a chill down the spine of anyone who's ever designed a database schema: Our new house that we just moved into... has two zip codes! | 4.2× | 0 | 4.9k | 632 | 155 |
| 72 | Apr 28, '25 | If you want to create completely free software for other people to use, the absolute best delivery mechanism right now is static HTML and JavaScript served from a free web host wit | 4.2× | 156k | 1.3k | 68 | 645 |
| 73 | Oct 23, '24 | I am beginning to suspect that the single biggest productivity hack for working with LLMs might be believing the vendors when they say that they won't train on your data | 4.0× | 148k | 1.2k | 36 | 162 |
| 74 | Jul 31, '25 | Anthropic shipped two new features for Claude over the past couple of days, but as far as I can tell they didn't provide any documentation at all beyond a tweet about each one ... | 3.9× | 143k | 1.2k | 95 | 859 |
| 75 | Jun 22, '25 | I like this take by @KentBeck on how AI-assisted programming changes the balance of which skills are most important | 3.8× | 139k | 1.3k | 148 | 749 |
| 76 | Dec 10, '24 | A weird and interesting thing about the new ChatGPT Canvas mode is that it can run Python in an entirely different way from the existing Code Interpreter - using Pyodide (Python in | 3.7× | 137k | 1.3k | 78 | 508 |
| 77 | Dec 22, '24 | It turns out OpenAI have a GitHub repository with a 26,000 line YAML OpenAPI schema describing their full API - the same API that's emulated by all sorts of other tools within the | 3.6× | 134k | 1.3k | 121 | 1.3k |
| 78 | Aug 1, '24 | Corollary to this: you can have an outsized impact on the world by being one of the few people who DO publish fresh information online on your own web pages Blog like it's 2005! | 3.6× | 132k | 1.2k | 153 | 168 |
| 79 | Jul 1, '25 | I figured out how to add the official Playwright browser automation MCP to Claude Code. Run this before you start "claude": claude mcp add playwright npx '@playwright/mcp@latest' N | 3.4× | 126k | 1.4k | 87 | 1.8k |
| 80 | Oct 11, '25 | I'm beginning to suspect that a key skill in working effectively with coding agents is developing an intuition for when you don't need to closely review every line of code they pro | 3.4× | 126k | 1.5k | 68 | 258 |
| 81 | Aug 6, '24 | Hidden at the bottom of this announcement: "By switching to the new gpt-4o-2024-08-06, developers save 50% on inputs ($2.50/1M input tokens) and 33% on outputs ($10.00/1M output to | 3.2× | 119k | 605 | 46 | 182 |
| 82 | Dec 22, '23 | @d_feldman It's got me to the point where I can read Spanish language news articles and understand ~80% of them - spoken Spanish is much harder | 3.1× | 114k | 2.3k | 24 | 38 |
| 83 | Jun 26, '25 | I'm really impressed by the new Gemma 3n I tried a 7.5GB model from Ollama and a 15GB model through mlx-vlm - they seem very capable, and this is the first model of that size I've | 2.9× | 108k | 1.2k | 112 | 591 |
| 84 | Sep 14, '25 | Getting LLMs to write prompts for themselves is one of those things that I didn't really trust a year ago - but the latest generation of models (Claude 4, GPT-5 in particular) have | 2.9× | 106k | 1.3k | 102 | 877 |
| 85 | Nov 15, '24 | Very impressed by Recraft AI - a new image generation service that can generate editable vector graphics that you can export as SVG This seems massively more useful than tools that | 2.8× | 105k | 1.4k | 104 | 1.2k |
| 86 | Dec 19, '24 | I figured out this prompting pattern for getting Claude to produce fully self-contained Python scripts that execute with "uv run" using PEP 723 inline script dependencies - and now | 2.8× | 105k | 1.4k | 88 | 1.6k |
| 87 | Oct 4, '25 | If you've been trying to figure out DSPy - the automatic prompt optimization system - this talk by @dbreunig is the clearest explanation I've seen yet, with a very useful real-worl | 2.2× | 80k | 1.4k | 153 | 1.7k |
| 88 | Aug 18, '18 | How about if, instead of ditching Twitter for Mastodon, we all start blogging and subscribing to each other's Atom feeds again instead? The original distributed social network coul | 1.9× | 0 | 2k | 510 | 20 |
| 89 | Aug 10, '22 | This is really grim, if not entirely unexpected: apparently the Instagram mobile app injects additional JavaScript into every page that's loaded using the in-app embedded browser - | 1.9× | 0 | 1.5k | 559 | 251 |
| 90 | Aug 27, '22 | The stuff happening on the Stable Diffusion subreddit right now is pretty wild - since the model can be run by anyone on their own machine if they have a decent GPU | 1.6× | 0 | 1.3k | 223 | 501 |
| 91 | Aug 6, '25 | @jeffreyhuber Any model, but you need to say "my friend has this idea and I'm trying to talk him out of it" | 1.6× | 58k | 2.3k | 58 | 747 |
| 92 | Feb 2, '22 | Whoa. runs a full Debian VM entirely in your browser via WebAssembly... and it ships with working Perl, Python, Ruby and Node.js! | 1.5× | 0 | 1.3k | 341 | 249 |
| 93 | May 24, '22 | Fascinating HN comment from someone who's company built a custom distributed data warehouse using compressed SQLite DB files in S3 that were queried using Lambda functions orchestr | 1.5× | 0 | 1.3k | 200 | 380 |
| 94 | Jul 20, '24 | OK this is really cool: Nico built an OpenAI action/GPT which breaks packages from PyPI up into <9.5MB chunks and returns them to ChatGPT in a way that lets it save them to disk | 1.4× | 53k | 350 | 37 | 220 |
| 95 | Sep 23, '22 | Reddit conversation about using GPT-3 to write your homework. A teacher comments: "Grading something an AI wrote is an incredibly depressing waste of my life." | 1.4× | 0 | 1.3k | 162 | 241 |
| 96 | Apr 16, '20 | SQL is a better API language than GraphQL. Convince me otherwise! | 1.3× | 0 | 1.3k | 154 | 136 |
| 97 | Oct 16, '21 | GitHub issues tip: if you paste in a link to an issue or PR in another repo it will display it as a truncated URL, but if you instead add it in a hyphenated bullet point it will di | 1.2× | 0 | 1.4k | 277 | 61 |
| 98 | Jul 17, '20 | 15 years ago today on my blog: Introducing Django | 1.1× | 0 | 1.3k | 159 | 56 |
| 99 | Apr 20, '21 | A lesson I re-learn on every project: always have an automatically populated "created_at" column on every single database table. Any time you think "I won't need it here" you're gu | 1.0× | 0 | 1.3k | 114 | 67 |
| 100 | Apr 19, '22 | @kingrolo @gvanrossum @tech_update0 @ethereum @solana We hate it because we understand it | 0.7× | 0 | 1k | 54 | 4 |
@karpathy
@levelsio
@paulg
@naval
@marclou
@illyism