Random branches are collected, scanned to 3D, and connected with servos. Then a neural network is trained to control this "robot".

Just read some of the samples of text generated with their neural networks, insane.

"Pérez and his friends were astonished to see the unicorn herd. These creatures could be seen from the air without having to move too much to see them - they were so close they could touch their horns. While examining these bizarre creatures the scientists discovered that the creatures also spoke some fairly regular English. Pérez stated, "We can see, for example, that they have a common 'language,' something like a dialect or dialectic."

They released a smaller model, I got it running on Sandy. It's fairly straight forward: https://github.com/openai/gpt-2

Really nice visualizations for rectifier (ReLU) neural nets, illustrating the effects of skip-connections, depth, width, etc. on the loss function curvature.

"We construct a novel large-scale, multi-modal dataset by pairing geo-referenced Wikipedia articles with satellite imagery of their corresponding locations."

Prediction-based method for encouraging reinforcement learning agents to explore their environments through curiosity (reward for unfamiliar states). Learns some games without any extrinsic reward!

Fun failure case: agent gets stuck in front of TV.

"Without the relevant data for reasoning about the state of AI technology, we are essentially "flying blind" in our conversations and decision-making related to AI [...] The AI Index is an initiative to track, collate, distill and visualize data relating to artificial intelligence. "