AI is incredibly consequential for our society. Here's what I read to help me think about it.

The Last Question

AI Code is Hollowing Out Open Source, and Maintainers are Looking the Other Way

AI’s Unpaid Debt: How LLM Scrapers Destroy the Social Contract of Open Source

The Future of Everything is Lies, I Guess

Anthropic’s philosopher answers your questions

OSGym: Scalable OS Infra for Computer Use Agents

On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜

Why the AI world is suddenly obsessed with a 160-year-old economics paradox

AI may be making us think and write more alike

Good Taste the Only Real Moat Left

Sam Altman May Control Our Future—Can He Be Trusted?

Is Copyright Law the Wrong Weapon Against AI?

Today is my last day at Anthropic. I resigned. Here is the letter I shared with my colleagues, explaining my decision.

You Will Know Nothing and You Will Be Happy

The Floor Is the Ceiling

#494 – Jensen Huang: NVIDIA - The $4 Trillion Company & the AI Revolution | Lex Fridman Podcast

Open Source AI is the Path Forward

Personal Superintelligence

GPT is revolutionary

Einstein–Szilard letter

Machines of Loving Grace

The Urgency of Interpretability

The Adolescence of Technology

Illustrating Reinforcement Learning from Human Feedback (RLHF)

Don’t trust AI to talk accurately about itself: Bard wasn’t trained on Gmail

CNET's Article-Writing AI Is Already Publishing Very Dumb Errors

Making ChatGPT Useful

GPT4 should be part of your toolkit

Building A Virtual Machine inside ChatGPT

Programming AIs worry me

Everything is Ghibli

The diminished art of coding

AI ambivalence

My AI Skeptic Friends Are All Nuts

Attention Is All You Need

AI Changes Everything

I Think AI Would Kill my Wife

Why I stopped using AI code editors

Speed matters

ChatGPT: it makes things up (that's its job)

ChatGPT: abstract logic and the doubling down bias

ChatGPT: it has no inner monologue or meta-cognition

Don’t Let Architecture Astronauts Scare You

Yes, and...

Why I’m not letting the juniors use GenAI for coding

Lena

Scaling Laws for Neural Language Models

Is there a no-AI audience?

Three Observations

Abundant Intelligence

The Gentle Singularity

Darwin Among the Machines

On DeepSeek and Export Controls

Thoughts on slowing the fuck down

Why the future doesn't need us.

Ethan's AI Reading List

Here are some of the papers we read in CS 2620.

How do peer-to-peer programs discover each other?

IPFS - Content Addressed, Versioned, P2P File System

Multi-agentic Software Development is a Distributed Systems Problem (AGI can't save you from it)

Viewstamped Replication Revisited

Small Is Not Always Beautiful

Clustering and Sharing Incentives in BitTorrent Systems

An empirical study of the reliability of UNIX utilities

End-to-end arguments in system design

Consensus in the presence of partial synchrony

Millions of Tiny Databases

Brewer's conjecture and the feasibility of consistent, available, partition-tolerant web services

The Google file system

The Google File System

There is more consensus in Egalitarian parliaments

Spanner: Google’s Globally-Distributed Database

The Chubby lock service for loosely-coupled distributed systems

The part-time parliament

Frangipani: a scalable distributed file system

Viewstamped Replication: A New Primary Copy Method to Support Highly-Available Distributed Systems

Another advantage of free choice (Extended Abstract): Completely asynchronous agreement protocols

Paxos vs Raft: have we reached consensus on distributed consensus?

Paxos made live: an engineering perspective

Paxos Made Simple

The weakest failure detector for solving consensus

Unreliable failure detectors for reliable distributed systems

A hundred impossibility proofs for distributed computing

Impossibility of distributed consensus with one faulty process

In Search of an Understandable Consensus Algorithm

CS 2620

Here are some of the papers we read in CS 1610.

Simultaneous Multithreading: A Platform for Next-Generation Processors

The Turtles Project: Design and Implementation of Nested Virtualization

Unikernels: library operating systems for the cloud

Unikernels: the rise of the virtual library operating system

A fork() in the road

An analysis of Linux scalability to many cores

The multikernel: a new OS architecture for scalable multicore systems

A case against (most) context switches

Are virtual-machine monitors microkernels done right?

Are virtual machine monitors microkernels done right?

seL4: formal verification of an operating-system kernel

Lessons Learned from 30 Years of MINIX – Communications of the ACM

Open Sources: Voices from the Open Source Revolution

The nucleus of a multiprogramming system

Practical Safe Linux Kernel Extensibility

How to read a paper

How to read a research paper

Memory resource management in VMware ESX server

CS 1610

Here are some of the papers we read in the HSRG.

Harvard Systems Reading Group

I'm a hobbyist philosophy reader, here are some papers I have read.

Affinity, Worth, and Fecundity: On Susan Wolf’s Advice for Living a Meaningful Life

Famine, Affluence, and Morality

Moral Saints

80,000 Hours for the Common Good

Against ‘Effective Altruism’

Philosophy Readings

The rise of AI has made me appreciate the importance of deep expertise even more. Here’s what I’m reading to learn.

Distributed Compression in the Era of Machine Learning: A Review of Recent Advances

Creating a vDSO: the Colonel's Other Chicken | Linux Journal

__structuredAttrs in Nix

East End Functions : Level of Indirection

a = b = c, a strange consequence of operator associativity – Belay the C++

[History of C++] Explanation on why the keyword `class` has no more reason to exist – Belay the C++

Welcome to my virtual library.

Libraries:

Ethan's AI Reading List

CS 2620

CS 1610

Harvard Systems Reading Group

Philosophy Readings

Interesting Technical Readings