How to Give a Good Talk
In computer science, conferences are a focal point of academic attention. Conferences are a moment where computing communities—distributed over the globe—come together. Giving a talk at a conferenc…
The latest blog posts to take off on Hacker News. 80,832 posts.
In computer science, conferences are a focal point of academic attention. Conferences are a moment where computing communities—distributed over the globe—come together. Giving a talk at a conferenc…
A quick note on the importance of reading the terms of service!
A new study co-authored by a University of Wisconsin–Madison professor finds that life expectancy gains made by high-income countries in the first half of the 20th century have slowed significantly, and that none of the generations born after 1939 will reach 100 years of age on average. Published in the journal…
Americans are having a record low amount of sex, especially young adults.
You see a series of checkboxes checked. Schedules met. Requirements satisfied. Demos delivered. It's a good day. Good job, you, good job! A promotion is in sight.
Neural networks are more useful when placed in a suitable, specialized environment.
In which I attempt to be pragmatic. Are you allowed to run whatever computer program you want on the hardware you own? This is a question where freedom, practicality, and reality all collide into a mess. Google has recently announced that Android users will only be able to install apps which have been digitally signed…
Analytics are a military and financial tool. Your blog is not and you shouldn’t surveil you readers.
A checklist for configuring Mozilla Firefox for a more private browsing experience.
From Multi-Head to Latent Attention: The Evolution of Attention Mechanisms What is attention? In any autoregressive model, the prediction of the future tokens is based on some preceding context …
SynthID is a tool to watermark and identify AI-generated content, helping to foster transparency and trust in generative AI.
Writing a Hypervisor in 1,000 Lines
When designing software systems, do the simplest thing that could possibly work. It’s surprising how far you can take this piece of advice. I genuinely think…
The official JS client for connecting to ClickHouse.
Is SQLite durable by default? What settings guarantee durability? The documentation and even comments from its creator give conflicting answers.
Do you register with Google, Amazon or Microsoft to use the web?
Building Marco meant committing to IMAP and cross-platform support. Here is why offline-first proved far harder than we expected, and what we learned.
W. Jherek Swanger writes in his introduction to Camillo 7 Agrippa's "A Treatise on the Science of Arms, with a philosophical Dialogue (1553)" that:
DeepSeek is a popular open-source large language model (LLM) praised for its strong performance. However, its large size and unique architecture, which uses Multi-head Latent Attention (MLA) and Mixture of Experts (MoE), require an advanced system for efficient serving at scale.
I didn't misclick any buttons. My best wasn't good enough. I'm not good enough.
We're thrilled to introduce grok-code-fast-1, a speedy and economical reasoning model that excels at agentic coding.
A deep dive into the convoluted and fascinating story of one of the most important keys on the keyboard
Claude Free, Pro, and Max users can now opt in to having conversations used to improve models, with a five-year retention period for participants.
Since I love collecting questionable analogies for LLMs, here's a new one I just came up with: an LLM is a lossy encyclopedia. They have a huge array of facts …
In the summer of 2022 I built a C64-based theremin, described and explained here . The theremin, of course, is one of the oldest electronic instruments. Its main drawback—and strength—is that it's incredibly hard to master. The performer has ultimate control over volume and pitch, but it takes months of practice…
Evan briefly discusses rss
I've been using Nim for about 1-2 years now, and I believe the language is undervalued. It's not perfect, of course, but it's pleasant to write and read. My personal website uses Nim .
TL;DR: all the assumptions you have about processor design are wrong and if you are unlucky you will never run into problems that users do through sheer chance.
Together, the Bitter Lesson and Scaling Laws reveal that the god of Compute we worship is yoked to an even greater one — the god of Data.