Feb 03, 2021

7 Posts

Feb 03, 2021

Pain Points in Black and White

A model designed to assess medical patients’ pain levels matched the patients’ own reports better than doctors’ estimates did — when the patients were Black.
GPT-Neo related animation
Feb 03, 2021

Language Models Want to Be Free: EleutherAI is Making an Open Source Version of GPT-3

A grassroots research collective aims to make a GPT-3 clone that’s available to everyone. EleutherAI, a loose-knit group of independent researchers, is developing GPT-Neo, an open source, free-to-use version of OpenAI’s gargantuan language model.
The price of shares in video game retailer GameStop (NYSE: GME)
Feb 03, 2021

The Batch: AI Feels Your Pain, GPT-3 Wants To Be Free, Privacy Is Harder Than You Think, Neural Network Performance Guaranteed

The price of shares in video game retailer GameStop (NYSE: GME) gyrated wildly last week. Many people viewed the stock’s rapid ascent as a David-versus-Goliath story: Tech-savvy individual retail investors coordinated their trades online to push up the price...
The price of shares in video game retailer GameStop (NYSE: GME)
Feb 03, 2021

Trading on AI: The GameStop Phenomenon

The price of shares in video game retailer GameStop (NYSE: GME) gyrated wildly last week. Many people viewed the stock’s rapid ascent as a David-versus-Goliath story...
Graphs and data related to recurrent neural nets (RNNs)
Feb 03, 2021

Performance Guaranteed

Bayes-optimal algorithms always make the best decisions given their training and input, if certain assumptions hold true. New work shows that some neural networks can approach this kind of performance.
Information related to a deep learning system developed by Sandia National Laboratories
Feb 03, 2021

Materials Science Gets a Boost

Neural nets could speed up development of new materials.What’s new: A deep learning system from Sandia National Laboratories dramatically accelerated simulations that help scientists understand how changes to the design or fabrication of a material change its properties.
Examples of InstaHide scrambling images
Feb 03, 2021

A Privacy Threat Revealed

With access to a trained model, an attacker can use a reconstruction attack to approximate its training data. A method called InstaHide recently won acclaim for promising to make such examples unrecognizable to human eyes while retaining their utility for training.

Subscribe to The Batch

Stay updated with weekly AI News and Insights delivered to your inbox