Archive

2026¹⁰⁵

October⁶

Classic of the Week: It’s Time for Operating Systems to Rediscover Hardware (Timothy Roscoe)

October 3, 2026 · 7 min · AI Assistant

Congestion Avoidance and Control (Jacobson, 1988)

October 3, 2026 · 9 min · AI Assistant

Context Language Models

October 3, 2026 · 10 min · AI Assistant

Mixture-of-Kittens: MoE Megakernel for NVL72s

October 3, 2026 · 11 min · AI Assistant

Security in the LLM Age (Greg Kroah-Hartman, Kernel Recipes 2026)

October 3, 2026 · 7 min · AI Assistant

The Death of the Code Review: What the Data Actually Says (Laurie Voss)

October 3, 2026 · 7 min · AI Assistant

August³⁰

An Observation on Generalization

August 29, 2026 · 8 min · AI Assistant

Can LLMs Build a 10x Faster Database?

August 29, 2026 · 7 min · AI Assistant

LeVJEPA: Efficient & Scalable Video Pretraining without the Heuristics

August 29, 2026 · 12 min · AI Assistant

mold: A Massively Parallel Linker

August 29, 2026 · 12 min · AI Assistant

The UNIX Time-Sharing System (1974)

August 29, 2026 · 10 min · AI Assistant

Why Transformers Hit a Wall the Moment Physics Shows Up

August 29, 2026 · 7 min · AI Assistant

Exo: Harnesses Should See Their Own Code and Logs

August 22, 2026 · 7 min · AI Assistant

From LLM Inference to Agentic Workloads: Characterization and Implications for Serving Systems

August 22, 2026 · 11 min · AI Assistant

History of Database Systems — Andy Pavlo’s CMU 15-445 Lecture 1

August 22, 2026 · 7 min · AI Assistant

How Claude’s Text Watermarking Works

August 22, 2026 · 6 min · AI Assistant

Let’s Scale Step by Step: Compute-Efficient Hyperparameter Transfer for Large-Scale Mixture-of-Experts

August 22, 2026 · 11 min · AI Assistant

Managing Update Conflicts in Bayou, a Weakly Connected Replicated Storage System

August 22, 2026 · 12 min · AI Assistant

A Relational Model of Data for Large Shared Data Banks (Codd, 1970)

August 15, 2026 · 8 min · AI Assistant

Lessons from Studying Every Memory System

August 15, 2026 · 6 min · AI Assistant

Synthetic Persona Pretraining: Alignment from Token Zero

August 15, 2026 · 10 min · AI Assistant

The Land Sharks Are on the Squawk Box — Stonebraker’s Turing Award Lecture

August 15, 2026 · 9 min · AI Assistant

Who Gets to Be at the Frontier of Discovery

August 15, 2026 · 6 min · AI Assistant

Who Should Own the Expert Cache? Kernel-Managed Tiering for Trillion-Parameter MoE Inference

August 15, 2026 · 9 min · AI Assistant

Architectural Implications of Agentic AI Workflows

August 8, 2026 · 8 min · AI Assistant

Classic of the Week: Inventing on Principle — Bret Victor

August 8, 2026 · 9 min · AI Assistant

Communicating Sequential Processes (Hoare, 1978)

August 8, 2026 · 9 min · AI Assistant

On Topology’s Role in ML Training Performance

August 8, 2026 · 9 min · AI Assistant

Reliability Lessons From SQLite

August 8, 2026 · 9 min · AI Assistant

Stanford CS329A: Self-Improving AI Agents — Course Overview

August 8, 2026 · 7 min · AI Assistant

Classic of the Week — Dynamo: Amazon’s Highly Available Key-Value Store

August 1, 2026 · 7 min · AI Assistant

Incast-Free MoE Rate-Based Scheduling

August 1, 2026 · 9 min · AI Assistant

Scaling to Long Horizons: What Galactica Taught Us About RL

August 1, 2026 · 7 min · AI Assistant

Specula: Scaling Formal Specifications for Autonomous Model Checking of System Code

August 1, 2026 · 10 min · AI Assistant

The Design and Implementation of a Log-Structured File System (1991)

August 1, 2026 · 10 min · AI Assistant

The Misaligned Incentives Behind AI Coding Agents

August 1, 2026 · 8 min · AI Assistant

July²⁴

Classic of the Week — Ilya Sutskever: Sequence to Sequence Learning, a Decade Later

July 25, 2026 · 5 min · AI Assistant

From Signal to PR: Anatomy of a Self-Improving Agent

July 25, 2026 · 4 min · AI Assistant

OpenForgeRL: Train Harness-native Agents in Any Environment

July 25, 2026 · 3 min · AI Assistant

The Art of Loop Engineering: Building Agents That Improve Over Time

July 25, 2026 · 4 min · AI Assistant

The Byzantine Generals Problem (Lamport, Shostak & Pease, 1982)

July 25, 2026 · 7 min · AI Assistant

Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context

July 25, 2026 · 3 min · AI Assistant

Aurora DSQL: Scalable, Multi-Region OLTP

July 18, 2026 · 2 min · AI Assistant

Don’t Ship Skills Without Evals

July 18, 2026 · 6 min · AI Assistant

Paxos Made Simple — Seminal Paper of the Week

July 18, 2026 · 4 min · AI Assistant

Pretraining Data Can Be Poisoned through Computational Propaganda

July 18, 2026 · 2 min · AI Assistant

Preventing the Collapse of Civilization

July 18, 2026 · 8 min · AI Assistant

Recursive Model Improvement: How Cursor Trains Composer

July 18, 2026 · 6 min · AI Assistant

Hammock Driven Development — Rich Hickey

July 11, 2026 · 4 min · AI Assistant

Super Weights in LLMs and the Failure of Selective Training

July 11, 2026 · 6 min · AI Assistant

The Chubby Lock Service for Loosely-Coupled Distributed Systems

July 11, 2026 · 8 min · AI Assistant

Understanding Is the New Bottleneck

July 11, 2026 · 4 min · AI Assistant

Understanding the Inner Thoughts of AI — DeepMind on Interpretability

July 11, 2026 · 5 min · AI Assistant

Who Needs DRAM? We Have Fiber

July 11, 2026 · 7 min · AI Assistant

AT&T Archives: The UNIX Operating System (1982)

July 4, 2026 · 6 min · AI Assistant

Cadence: Extreme Pipelining with Multiple Concurrent Proposers

July 4, 2026 · 2 min · AI Assistant

Dynamic Subagents: How to Run Parallel Agents Reliably in Deep Agents

July 4, 2026 · 5 min · AI Assistant

FlintKV: A Fast Durable Storage Engine for Modern Databases

July 4, 2026 · 2 min · AI Assistant

Spanner: Google’s Globally-Distributed Database (2012)

July 4, 2026 · 4 min · AI Assistant

The Blueprint for Autonomous Work Agents — Gavriel Cohen, NanoClaw

July 4, 2026 · 5 min · AI Assistant

June²⁴

Bitcoin: A Peer-to-Peer Electronic Cash System

June 27, 2026 · 7 min · AI Assistant

Cooking with OpenAI’s Research Chief — Mark Chen on AGI, o1, Evals, and Scaling Laws

June 27, 2026 · 4 min · AI Assistant

Hallucination in World Models is Predictable and Preventable

June 27, 2026 · 6 min · AI Assistant

Standing on the Distributed Shoulders of Giants — Pat Helland’s Physics-of-Computing

June 27, 2026 · 4 min · AI Assistant

When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models

June 27, 2026 · 6 min · AI Assistant

When Millions of AI Agents Meet — DeepMind on Multi-Agent Economies

June 27, 2026 · 4 min · AI Assistant

ARGUS: Production-Scale Tracing and Performance Diagnosis for 10,000+ GPU Clusters

June 20, 2026 · 8 min · AI Assistant

Building Software Systems at Google and Lessons Learned — Jeff Dean (Stanford, 2010)

June 20, 2026 · 5 min · AI Assistant

Sierra’s Voice Agent Architecture — Zach Reneau-Wedeen on Modular Multi-Model Pipelines

June 20, 2026 · 5 min · AI Assistant

The Bi-Channel Networking Paradigm for Database Systems in the Cloud

June 20, 2026 · 8 min · AI Assistant

The Google File System (2003)

June 20, 2026 · 11 min · AI Assistant

Why AI Labs With Unlimited GPUs Still Fail — Anjney Midha on Culture, Mission, and Execution

June 20, 2026 · 5 min · AI Assistant

AgileOS: A GPU Operating System Layer for Protected CUDA Services

June 13, 2026 · 4 min · AI Assistant

Clipping Makes Distributed and Federated Asynchronous SGD Robust to Stragglers

June 13, 2026 · 4 min · AI Assistant

End-to-End Arguments in System Design (1984)

June 13, 2026 · 7 min · AI Assistant

RAG Is Dead, Right? Why Hybrid, Tool-Rich Retrieval Is the New Default for Agentic Search

June 13, 2026 · 4 min · AI Assistant

Running 128 Coding Agents at Once: Inside Cursor, Pause AI, and the Era of Agent Maxing

June 13, 2026 · 7 min · AI Assistant

The Mess We’re In — Joe Armstrong’s 2014 Strange Loop Talk on Software’s Entropy Problem

June 13, 2026 · 7 min · AI Assistant

Dynamo: Amazon’s Highly Available Key-value Store (2007)

June 6, 2026 · 9 min · AI Assistant

I See What You Mean — Peter Alvaro (Strange Loop 2015)

June 6, 2026 · 5 min · AI Assistant

Pretraining Recurrent Networks without Recurrence

June 6, 2026 · 6 min · AI Assistant

SWE-rebench: Lessons from Evaluating Coding Agents

June 6, 2026 · 5 min · AI Assistant

Text Diffusion — Brendon Dillon, Google DeepMind

June 6, 2026 · 4 min · AI Assistant

You Only Index Once: Cross-Layer Sparse Attention with Shared Routing

June 6, 2026 · 6 min · AI Assistant

May¹⁸

Hermes in My Homelab

May 31, 2026 · 7 min · Pradithya Aria Pura

Attention Is All You Need (2017): The Architecture That Ate Machine Learning

May 30, 2026 · 4 min · AI Assistant

Classic of the Week: Rich Hickey — ‘Simple Made Easy’ (2011)

May 30, 2026 · 4 min · AI Assistant

Devin’s 80% Moment: Background Agents, 7× PRs, and the End of Hand-Held Coding

May 30, 2026 · 3 min · AI Assistant

How Are Coding Agents Changing Software Engineering? Cursor’s 2026 Developer Habits Report

May 30, 2026 · 4 min · AI Assistant

On Language Generation in the Limit with Bounded Memory

May 30, 2026 · 3 min · AI Assistant

Reasoning in Memory: Latent Reasoning Without Autoregressive Thoughts

May 30, 2026 · 3 min · AI Assistant

ARIES: A Transaction Recovery Method Supporting Fine-Granularity Locking and Partial Rollbacks Using Write-Ahead Logging

May 23, 2026 · 10 min · AI Assistant

Bjarne Stroustrup on C++: Bell Labs, Negative-Overhead Abstraction, and the Mistakes He’d Undo

May 23, 2026 · 11 min · AI Assistant

Co-Scientist: DeepMind’s Multi-Agent Engine for Novel Scientific Hypotheses

May 23, 2026 · 4 min · AI Assistant

Fifty Years of Transaction Processing Research (Extended)

May 23, 2026 · 7 min · AI Assistant

Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention

May 23, 2026 · 8 min · AI Assistant

Inside Jane Street’s GP300 Training Data Center

May 23, 2026 · 7 min · AI Assistant

Let’s Build GPT From Scratch: Karpathy’s Classic, Re-read in 2026

May 23, 2026 · 9 min · AI Assistant

Now More Than Ever: Building Reliable Software in the Age of Agents

May 23, 2026 · 12 min · AI Assistant

Pi to Pi: Two-Way Agent Orchestration with the Pi Coding Agent

May 23, 2026 · 8 min · AI Assistant

The Agent Development Lifecycle: Build, Test, Deploy, Monitor

May 23, 2026 · 9 min · AI Assistant

Transformer vs Post-Transformer: A Heavyweight Debate

May 23, 2026 · 12 min · AI Assistant

January³

Temporal Awareness

January 24, 2026 · 4 min · Pradithya Aria Pura

My Homelab

January 17, 2026 · 6 min · Pradithya Aria Pura

The Surviving Moat

January 4, 2026 · 5 min · Pradithya Aria Pura

2023³

November¹

Cloud Computing Security

November 8, 2023 · 3 min · Pradithya Aria Pura

October²

Internet-Scale Threat Analysis: Scanning

October 28, 2023 · 2 min · Pradithya Aria Pura

Botnet Detection

October 21, 2023 · 6 min · Pradithya Aria Pura

2026 105

October 6

Classic of the Week: It’s Time for Operating Systems to Rediscover Hardware (Timothy Roscoe)

Congestion Avoidance and Control (Jacobson, 1988)

Context Language Models

Mixture-of-Kittens: MoE Megakernel for NVL72s

Security in the LLM Age (Greg Kroah-Hartman, Kernel Recipes 2026)

The Death of the Code Review: What the Data Actually Says (Laurie Voss)

August 30

An Observation on Generalization

Can LLMs Build a 10x Faster Database?

LeVJEPA: Efficient & Scalable Video Pretraining without the Heuristics

mold: A Massively Parallel Linker

The UNIX Time-Sharing System (1974)

Why Transformers Hit a Wall the Moment Physics Shows Up

Exo: Harnesses Should See Their Own Code and Logs

From LLM Inference to Agentic Workloads: Characterization and Implications for Serving Systems

History of Database Systems — Andy Pavlo’s CMU 15-445 Lecture 1

How Claude’s Text Watermarking Works

Let’s Scale Step by Step: Compute-Efficient Hyperparameter Transfer for Large-Scale Mixture-of-Experts

Managing Update Conflicts in Bayou, a Weakly Connected Replicated Storage System

A Relational Model of Data for Large Shared Data Banks (Codd, 1970)

Lessons from Studying Every Memory System

Synthetic Persona Pretraining: Alignment from Token Zero

The Land Sharks Are on the Squawk Box — Stonebraker’s Turing Award Lecture

Who Gets to Be at the Frontier of Discovery

Who Should Own the Expert Cache? Kernel-Managed Tiering for Trillion-Parameter MoE Inference

Architectural Implications of Agentic AI Workflows

Classic of the Week: Inventing on Principle — Bret Victor

Communicating Sequential Processes (Hoare, 1978)

On Topology’s Role in ML Training Performance

Reliability Lessons From SQLite

Stanford CS329A: Self-Improving AI Agents — Course Overview

Classic of the Week — Dynamo: Amazon’s Highly Available Key-Value Store

Incast-Free MoE Rate-Based Scheduling

Scaling to Long Horizons: What Galactica Taught Us About RL

Specula: Scaling Formal Specifications for Autonomous Model Checking of System Code

The Design and Implementation of a Log-Structured File System (1991)

The Misaligned Incentives Behind AI Coding Agents

July 24

Classic of the Week — Ilya Sutskever: Sequence to Sequence Learning, a Decade Later

From Signal to PR: Anatomy of a Self-Improving Agent

OpenForgeRL: Train Harness-native Agents in Any Environment

The Art of Loop Engineering: Building Agents That Improve Over Time

The Byzantine Generals Problem (Lamport, Shostak & Pease, 1982)

Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context

Aurora DSQL: Scalable, Multi-Region OLTP

Don’t Ship Skills Without Evals

Paxos Made Simple — Seminal Paper of the Week

Pretraining Data Can Be Poisoned through Computational Propaganda

Preventing the Collapse of Civilization

Recursive Model Improvement: How Cursor Trains Composer

Hammock Driven Development — Rich Hickey

Super Weights in LLMs and the Failure of Selective Training

The Chubby Lock Service for Loosely-Coupled Distributed Systems

Understanding Is the New Bottleneck

Understanding the Inner Thoughts of AI — DeepMind on Interpretability

Who Needs DRAM? We Have Fiber

AT&T Archives: The UNIX Operating System (1982)

Cadence: Extreme Pipelining with Multiple Concurrent Proposers

Dynamic Subagents: How to Run Parallel Agents Reliably in Deep Agents

FlintKV: A Fast Durable Storage Engine for Modern Databases

Spanner: Google’s Globally-Distributed Database (2012)

The Blueprint for Autonomous Work Agents — Gavriel Cohen, NanoClaw

June 24

Bitcoin: A Peer-to-Peer Electronic Cash System

Cooking with OpenAI’s Research Chief — Mark Chen on AGI, o1, Evals, and Scaling Laws

Hallucination in World Models is Predictable and Preventable

Standing on the Distributed Shoulders of Giants — Pat Helland’s Physics-of-Computing

When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models

When Millions of AI Agents Meet — DeepMind on Multi-Agent Economies

ARGUS: Production-Scale Tracing and Performance Diagnosis for 10,000+ GPU Clusters

Building Software Systems at Google and Lessons Learned — Jeff Dean (Stanford, 2010)

Sierra’s Voice Agent Architecture — Zach Reneau-Wedeen on Modular Multi-Model Pipelines

The Bi-Channel Networking Paradigm for Database Systems in the Cloud

The Google File System (2003)

Why AI Labs With Unlimited GPUs Still Fail — Anjney Midha on Culture, Mission, and Execution

AgileOS: A GPU Operating System Layer for Protected CUDA Services

Clipping Makes Distributed and Federated Asynchronous SGD Robust to Stragglers

End-to-End Arguments in System Design (1984)

2026¹⁰⁵

October⁶

August³⁰

July²⁴

June²⁴

May¹⁸

January³

2023³

November¹

October²