Best of 2024

The blog posts Hacker News loved most in 2024. 3,433 posts.

Any topic
2591

Using a LLM to compress text

Introduction Large language models are trained on huge datasets of text to learn the relationships and contexts of words within larger documents. These relationships are what allows the model to generate text. Recently I've read concerns about LLMs being trained on copyrighted text and reproducing it. This got me…

2593

Choosing what to read

Like many people, I spend a lot of time reading. Since there is much more that I could read than there is time for, I have to make choices about what I should read. In this short post I’m going to outline what my reading choices have become and why. My aim isn’t to suggest that anyone else should make the same choices…

2603

Listening with LLM

Overview # This is the first part of many posts I am writing to consolidate learnings on how to finetune Large Language Models (LLMs) to process audio, with the eventual goal of being able to build and host a LLM able to describe human voices. I am motivated to gain hands-on experience tinkering LLMs so, as much as…

2609

Code review antipatterns

Code review seems like a great idea, right? Two developers looking at the same code means two chances to spot problems; it spreads understanding of the way the project is evolving; the reviewer can learn useful tricks from reading the author’s code in detail, or spot opportunities to teach the author a useful trick…