Posts that commenters keep linking to, ranked by how many different people have cited them in Hacker News comments. 422 posts.
Chinese AI lab Moonshot AI announced Kimi K3 this morning, describing it as their “most capable model to date, with 2.8 trillion parameters”. It’s currently available via their website and …
Identifying them as such only lets companies like OpenAI off the hook.
Writing code is cheap now - Agentic Engineering Patterns
I believe that language models aren’t world models. It’s a weak claim — I’m not saying they’re useless, or that we’re done milking them. It’s also a fuzzy-sounding claim — with its trillion weights, who can prove that there’s something an LLM isn't a model of? But I hope to make my claim clear and persuasive enough…
It's been just over two years and two months since ChatGPT launched, and in that time we've seen Large Language Models (LLMs) blossom from a novel concept into one of the most craven cons of the 21st century — a cynical bubble inflated by OpenAI CEO Sam Altman built
An image kernel is a small matrix used to apply effects like the ones you might find in Photoshop or Gimp, such as blurring, sharpening, outlining or embossing. They're also used in machine learning for 'feature extraction', a technique for determining the most important portions of an image. In this context the…
"Prompt Engineering" emerged from the growth of language models to describe the process of applying prompting to effectively extract information from language models, typically for use in real-world applications.
Hacky repo to see what the Copilot extension sends to the server
Artistic interpretations of the 54 Roman emperors of the Principate, created with AI and historical references.
How to run Claude Code with the `--dangerously-skip-permissions` flag in a safe way
Inside every AI is a bigger AI, trying to get out
Ever since I tried ChatGPT and GPT-3, everything else feels painfully dumb and useless: Siri, Alexa, Google Home and all other “smart”…
Fictional short story about Clippy & AI hard takeoff scenarios grounded in contemporary ML scaling, self-supervised learning, reinforcement learning, and meta-learning research literature.
Plus how American professors are fighting back against the AI onslaught, a backlash over AI models in Vogue, and more.
Rob Pike (that Rob Pike) is furious. Here’s a Bluesky link for if you have an account there and a link to it in my thread viewer if you don’t. …
There is an extremely important question about the near-future of AI that almost no-one is asking. We’ve all seen the graphs from METR showing that the length of tasks AI agents can perform has been growing exponentially over the last 7 years. While GPT-2 could only do software engineering tasks t
Tonight, I took the time actually to read DeepMind’s AlphaCode paper, and to work through the example contest problems provided, and understand how I would’ve solved those problems, and how A…
Building agents with LLM (large language model) as its core controller is a cool concept. Several proof-of-concepts demos, such as AutoGPT, GPT-Engineer and BabyAGI, serve as inspiring examples. The potentiality of LLM extends beyond generating well-written copies, stories, essays and programs; it can be framed as a…
Study Shows That Even Experienced Developers Dramatically Overestimate Gains
The State Space Model taking on Transformers
GPT-3 is indeed a large step forward for AI text-generation, but there are very many caveats with the popular demos and use cases.
I shared a controversial take the other day at an event and I decided to write it down in a longer format: I’m afraid AI won't give us a compressed 21st century.
Wolfram expands Notebooks, integrating LLM functionality into the new chat cell and Chat Notebook. Stephen Wolfram explains how chat-enabled and chat-driven versions work.
A deep-dive into machine learning embeddings.
Many new Transformer architecture improvements have been proposed since my last post on “The Transformer Family” about three years ago. Here I did a big refactoring and enrichment of that 2020 post — restructure the hierarchy of sections and improve many sections with more recent papers. Version 2.0 is a superset of…
This article was originally written in 2014, but has had some minor improvements in December 2017 and January 2018 when I received some useful feedback after the article had unexpectedly been posted on Hacker News in December 2017.
Summary : Update to my last post analyzing the gzip+knn paper. Some of the datasets have heavily contaminated train/test tests. I show some ways to significantly speed-up the gzip method. I present more results of my "fair" reimplementation .
