Generating Interactive Worlds with Text
we present a machine learning (ML) approach to creating a cohesive and interesting world built from elements of the text-based fantasy game
ML models were trained to play the game by mimicking the actions and dialogues of human players in fixed settings built by crowd-workers.
In contrast, in this work, we study models for assembling the game itself rather than agents that play it
Paper:
https://arxiv.org/pdf/1911.09194.pdf

On the Morality of Artificial Intelligence
Alexandra Luccioni, Yoshua Bengio : https://arxiv.org/abs/1912.11945
#Society #AIEthics #ArtificialIntelligence

"Finally, machine learning interprets gene regulation clearly"
By Brian Stallard, Cold Spring Harbor Laboratory : https://phys.org/news/2019-12-machine-gene.html
#ArtificialIntelligence #Genetics #MachineLearning

"Differentiable Convex Optimization Layers"
CVXPY creates powerful new PyTorch and TensorFlow layers
Agrawal et al.: https://locuslab.github.io/2019-10-28-cvxpylayers/
#PyTorch #TensorFlow #NeurIPS2019

deeptraffic: DeepTraffic is a deep reinforcement learning competition, part of the MIT Deep Learning series. https://github.com/lexfridman/deeptraffic

"Optuna: A Next-generation Hyperparameter Optimization Framework"
Akiba et al.: https://arxiv.org/abs/1907.10902
#ArtificialIntelligence #DataScience #MachineLearning

Neural-Symbolic Cognitive Reasoning
Authors: D'Avila Garcez, Artur S., Lamb, Luís C., Gabbay, Dov M - https://www.springer.com/gp/book/9783540732457

Deep learning model for breast cancer detection beats five full-time radiologists and previous SOTA models from NYU and MIT
Paper: https://arxiv.org/pdf/1912.11027.pdf

Deep learning model for breast cancer detection beats five full-time radiologists and previous SOTA models from NYU and MIT

Deep learning approach improves accuracy of screening mammography
https://arxiv.org/abs/1912.11027

This story was told by the American inventor/entrepreneur/scientist Daniel Hillis, who did pioneering work in the application of parallel computers to artificial intelligence.

Richard Feynman's son, Carl, was working as an undergraduate assistant to Hillis in a project they called "Connection Machine." Feynman, which was interested in computing since Los Alamos, was following the project closely. One day after having lunch with Hillis, Feynman agreed to work at his startup company in the summer. Shortly after the company was incorporated, Feynman showed up at the headquarters, saying, "Richard Feynman reporting for duty. OK, boss, what's my assignment?" Everybody in the company was surprised by the unexpected situation, and after some discussion, they gave Feynman the assignment of advising "on the application of parallel processing to scientific problems." Feynman said: "That sounds like a bunch of baloney. Give me something real to do." So they sent him out to buy office supplies, which he promptly did.

HOW SCIENTISTS ARE UPLOADING YOUR BRAIN PATREON
ABSTRACT: A thesis for achieving mind-transfer and/or consciousness-transfer from biological wetware to a neuromorphic computer substrate via connectomics.
METHODS:
1:17 Worm-Upload to Robot
2:32 Mouse-Upload by the Allen Institute
4:16 Human-Uploads and the Connectome Project by Harvard, MIT, Washington
8:20 Brain Connectome
10:58 Brain Metabolome
12:48 Brain Proteome 14:10 Brain Moleculome
16:25 Conclusions

Youtube: https://www.youtube.com/watch?v=EsA2ibE6ut0

Is Deep Learning the Future of Medical Decision Making?
https://thegradient.pub/is-deep-learning-the-future-of-medical-decision-making/

The AI field is in its infancy
A great definition of deep learning would allow the field to be more welcoming, to extend its reach to other fields and to illuminate the state of discourse. #AIDebate

Open-source chatbot project from Microsoft!

https://www.profillic.com/paper/arxiv:1912.08904

Macaw: A conversational bot that enables research for tasks such as document retrieval, question answering, recommendation, and structured data exploration