AI research | The Decoder

Matthias Bastian

Feb 12, 2023

Toolformer language model uses external tools on its own

Robot hands holding various tools popping out of a computer monitor on a desk, illustration, AI art

Maximilian Schreiner

Feb 12, 2023

Nvidia can precisely control computer characters using only language

Jonathan Kemper

Feb 11, 2023

Openjourney is Stable Diffusion trained with Midjourney images

Matthias Bastian

Feb 7, 2023

Gen-1: Stable Diffusion startup introduces generative AI for video

An animated figure is swimming under water. At the bottom left, you can see a blue version of this figure without textures in the small image. The texture large in the image, a dark color with gold dust and the rendered water, were added by the system via text command only.

Maximilian Schreiner

Feb 6, 2023

Google Dreamix: This AI video editor can animate your teddy bears

Some photos of a stuffed bear on the left and an AI-generated image of the same stuffed bear walking on the right.

Jonathan Kemper

Feb 5, 2023

AI image models generate duplicates from training material - study

A robot stands in front of a mirror and looks at itself, comic style.

Maximilian Schreiner

Feb 4, 2023

Deepmind Ada brings foundation models to reinforcement learning

A scerenshot of a computer generatedN environment, in the center is a blue dot, the AI agent Ada.

Maximilian Schreiner

Feb 2, 2023

Stochastic parrot or world model? How large language models learn

A parrot and a raven play a board game

Matthias Bastian

Jan 30, 2023

Stanford DetectGPT and GPTZeroX: New tools for AI text recognition

A comic of a teacher critically examining a document in the classroom. Behind him are students on their benches.

Maximilian Schreiner

Jan 29, 2023

RecolorNeRF is like a basic Photoshop for NeRFs

A nerf scene with a car, an interfache to the right of the scene allows the car to change color.