Ai Research

Explore recommended reads, summarized by AI

AI Capabilities and Limitations

Examinations of what AI systems can and cannot do, and debates over their effects on human work and thinking.

AI's Effects on Work and Cognition

Language Model Mechanics

Technical explorations of how large language models process information and produce output.

How Language Models Generate Text

Popular this week

Pop Alexandru

A New AI Paper Found a Problem With How Language Models Learn to Predict Multiple Tokens

3d ago

Kevin O'Shaughnessy

Top 30 AI Safety Researchers

Aug 3

Anna Ly

My BlueDot Technical AI Safety Participation Experience

May 20

AI Made Practical

AI Tool Steps

Which AI Is Best for Research? I Compared ChatGPT, Gemini, and Claude

Oct 24

Mubashir Burfat

I Tried AnswerThis for AI-Powered Research. Here’s What Actually Impressed Me

Sep 3

Selected by Medium

Code Like A Girl

Aleksandra Liutikova aka Java Senorita

The Pain Axis, in plain English: what happens when you’re rude to a chatbot

Sep 22

Daryle Serrant

The Hidden Structure Behind How AI Writes (Part 5): How Context Reshapes What Comes Next

Sep 12

Daryle Serrant

The Hidden Structure Behind How AI Writes (Part 2): Reverse Engineering AI Grammar

Jul 17

Antonio V. Franco

Why Studying Mathematics Will Always Be Necessary for Working in Computing

May 21

Latest stories

MayhemCode

MIT Let GPT-6 Edit Another AI’s Brain. Its Math Score Jumped Sixfold.

18h ago

AI Engineering

Anthropic Open-Sources FlashProve: The End of Unverified Model API Swaps

19h ago

Ankit Goel

An AI Just Edited Another AI’s Brain. No Training Required.

1d ago

GoPenAI

Satvik Jain

Weight-Space Programming Explained: LLMs That Edit Weights

1d ago

AIFix

gpt 6 astra weights: What MIT’s “Weight-Space Programming” Study Means for Small Businesses and…

2d ago

See more stories

Ai Research

AI Capabilities and Limitations

AI's Effects on Work and Cognition

Language Model Mechanics

How Language Models Generate Text

Popular this week

A New AI Paper Found a Problem With How Language Models Learn to Predict Multiple Tokens

Top 30 AI Safety Researchers

My BlueDot Technical AI Safety Participation Experience

Which AI Is Best for Research? I Compared ChatGPT, Gemini, and Claude

I Tried AnswerThis for AI-Powered Research. Here’s What Actually Impressed Me

Selected by Medium

The Pain Axis, in plain English: what happens when you’re rude to a chatbot

The Hidden Structure Behind How AI Writes (Part 5): How Context Reshapes What Comes Next

The Hidden Structure Behind How AI Writes (Part 2): Reverse Engineering AI Grammar

Why Studying Mathematics Will Always Be Necessary for Working in Computing

Latest stories

MIT Let GPT-6 Edit Another AI’s Brain. Its Math Score Jumped Sixfold.

Anthropic Open-Sources FlashProve: The End of Unverified Model API Swaps

An AI Just Edited Another AI’s Brain. No Training Required.

Weight-Space Programming Explained: LLMs That Edit Weights

gpt 6 astra weights: What MIT’s “Weight-Space Programming” Study Means for Small Businesses and…

Who to follow

Popular this week

A New AI Paper Found a Problem With How Language Models Learn to Predict Multiple Tokens

Top 30 AI Safety Researchers

My BlueDot Technical AI Safety Participation Experience

Which AI Is Best for Research? I Compared ChatGPT, Gemini, and Claude

I Tried AnswerThis for AI-Powered Research. Here’s What Actually Impressed Me

Selected by Medium

The Pain Axis, in plain English: what happens when you’re rude to a chatbot

The Hidden Structure Behind How AI Writes (Part 5): How Context Reshapes What Comes Next

The Hidden Structure Behind How AI Writes (Part 2): Reverse Engineering AI Grammar

Why Studying Mathematics Will Always Be Necessary for Working in Computing

Latest stories

MIT Let GPT-6 Edit Another AI’s Brain. Its Math Score Jumped Sixfold.

Anthropic Open-Sources FlashProve: The End of Unverified Model API Swaps

An AI Just Edited Another AI’s Brain. No Training Required.

Weight-Space Programming Explained: LLMs That Edit Weights

gpt 6 astra weights: What MIT’s “Weight-Space Programming” Study Means for Small Businesses and…

Who to follow