CapabilitiesIntermediate
Recursive Self-Improvement
What happens when an AI gets good at making AI better? The feedback loop that could turn steady progress into a sudden takeoff.
Read→5 min read
Science, beautifully explained
The ideas behind artificial general intelligence, made to be understood.
Long, visual, rigorous stories that take one concept at a time and give it the space and the craft it deserves.
The library
One concept per story. More going live over time.
InterpretabilityAdvanced
Mechanistic Interpretability
Reverse-engineering the circuits inside a neural network, neuron by neuron, to read what a model actually computes.
Soon
CapabilitiesIntro
Scaling Laws
Why bigger models, more data and more compute keep paying off on a startlingly smooth curve, and where that curve might bend.
Soon
AlignmentIntro
The Alignment Problem
Getting a powerful optimizer to want what we want turns out to be far harder than telling it what to do.
Soon
AlignmentIntermediate
Instrumental Convergence
Almost any goal, pursued hard enough, rewards the same sub-goals: gather resources, stay switched on, resist being changed.
Soon
AlignmentIntermediate
Reinforcement Learning from Human Feedback
How human thumbs-up and thumbs-down get distilled into a reward model that quietly shapes a chatbot's personality.
Soon