991
Recursive Language Models (RLMs)
We explore language models that recursively call themselves or other LLMs before providing a final answer. The aim is essentially unbounded context, and a way through context rot.
Top posts
Every blog post that took off on Hacker News, ranked by the votes of its best submission. 1,227 posts.
992
An overview of gradient descent optimization algorithms (2016)
Gradient descent is the preferred way to optimize neural networks and many other machine learning algorithms but is often used as a black box. This post explores how many of the most popular gradient-based optimization algorithms such as Momentum, Adagrad, and Adam actually work.
993
If you believe in "Artificial Intelligence", take five minutes to ask it
Here’s a Mastodon thread from a year ago. Just a quick check on how ChatGPT’s getting on … Me: Who reassigned the species Brachiosaurus brancai to its own genus, and when? ChatGPT…
994
Richard Stallman on ChatGPT
I define "intelligence" as being capable of knowing or understanding, at least within some domain. ChatGPT cannot know or understand anything, so it is not intelligence. It does not know what its output means. It has no idea that words can mean anything.
995
You can’t build a moat with AI (redux)
Using LLMs doesn't make a product special anymore
996
We reverse-engineered Flash Attention 4
Asynchrony, fast approximate exponents, and 10x more efficient softmax.
997
From text to token: How tokenization pipelines work
Understanding how search engines transform text into tokens through character filtering, tokenization, stemming, and stopword removal.
998
Using secondary school maths to demystify AI
Educators can show in secondary school maths that AI systems don’t think, making maths more interesting while teaching core concepts of AI.
999
Absolute Zero Reasoner
Current reasoning models trained with Reinforcement Learning with Verifiable Rewards (RLVR) often rely on manually curated datasets, raising scalability concerns and potentially limiting future AI growth beyond human-defined tasks. To address this, we propose the Absolute Zero paradigm, where a single model…
1000
Grok 3: Another win for the bitter lesson
Congratulations to the xAI team—and the advocates of the scaling laws
1001
The Unreliability of LLMs and What Lies Ahead
The Unreliability of LLMs & What Lies Ahead
1002
I tried coding with AI, I became lazy and stupid
Robots won't replace me, but another human probably will.
1003
1004
Android users can now use conversational editing in Google Photos
We’re expanding the ability to edit your images by simply asking in Google Photos to Android devices.
1005
The State of AI Coding Report 2025
Read the full Q2 2026 research report, explore the chart data, and download the PDF for free.
1006
1007
The Illusion of Thinking: Strengths and Limitations of Reasoning Models
Recent generations of frontier language models have introduced Large Reasoning Models (LRMs) that generate detailed thinking processes…
1008
It's OpenAI's world, we're just living in it
The best Stratechery content from the week of October 6, 2025, including OpenAI’s Windows strategy, Sam Altman and boundless ambition, and the future of creation.
1009
Three kinds of AI products work
The very first LLM-based product, ChatGPT, was just the ability to talk with the model itself: in other words, a pure chatbot. This is still the most popular…
1010
Critical vulnerability in LangChain – CVE-2025-68664
Check Point's AI Security Solutions deliver AI-driven threat prevention and unified security management for advanced protection across enterprise networks.
1011
Knowledge Management in the Age of AI
My personal website
1012
Time Series Forecasting with Graph Transformers
Introduction to Kumo Relational — the relational foundation model for predictive AI on relational data
1013
Phi-4 Reasoning Models
Microsoft continues to add to the conversation by unveiling its newest models, Phi-4-reasoning, Phi-4-reasoning-plus, and Phi-4-mini-reasoning. Learn more.
1014
1015
Tversky Neural Networks
Psychologically Plausible Deep Learning with Differentiable Tversky Similarity
1016
"ChatGPT said this" Is Lazy
When you paste ChatGPT’s response instead of your own feedback, you’re not being helpful. You’re being lazy and creating more work for everyone.
1017
If writing is thinking then what happens if AI is doing the writing and reading?
Then what happens if AI is doing the writing and reading?
1018
Deep Agents
Build deep agents that plan and execute complex tasks over longer time horizons. Learn the architecture behind Deep Research and Claude Code.
1019
Attention Wasn't All We Needed
Personal blog of Stephen Diehl - Software engineer writing about technology, programming, and the future
1020
The "AI 2027" Scenario: How realistic is it?
Everyone’s been talking about it. JD Vance read it. What does it actually tell us?