AI might be seemingly everywhere, but there are still plenty of things it can’t do – for now

Rethinking with Retrieval: Faithful Large Language Model Inference

Will we run out of data? An analysis of the limits of scaling datasets in Machine Learning

A simpler path to better computer vision

Evaluating Human-Language Model Interaction

Reconstructing Hand-Held Objects from Monocular Video

Undesirable biases in NLP: Averting a crisis of measurement

Solving brain dynamics gives rise to flexible machine-learning models

Not everything we call AI is actually ‘artificial intelligence’. Here’s what you need to know

On the Role of Parallel Data in Cross-lingual Transfer Learning