AI with AI
Episode 3.22: RIDE of the COV-all-cures
In COVID-related news, Andy and Dave discuss CloseLoop.ai and its release of an open-source toolkit for predicting people vulnerable to COVID-19. A Korean biotech company, Seegene, announces that it has used AI to create a coronavirus test. DarwinAI and research at the University of Waterloo announce COVID-Net, a convolutional neural network for detecting COVID-19 in chest x-rays. In non-COVID news, the White House releases its first annual report on AI. The U.S. intelligence community describes its interest in using explainable and interpretable AI. And Microsoft introduces a checklist that attempts to bridge the gap between the AI ethics community and ML practitioners. And House Science Committee members introduce the National AI Initiative Act, which aims to accelerate and coordinate federal investments in AI. In research, the NIH monitors brains replaying memories in real time, by examining neuron firing patterns for word pattern association (such as camel and lime). Facebook AI Research announces Rewarding Impact-Driven Exploration (RIDE), where agents are encouraged to take actions that have significant impact on the environment state. Researchers from the WHO and other institutions examine the landscape of AI applications to COVID-19. Andrea Gilli publishes The Brain and the Processor: Unpacking the Challenges of Human-Machine Interaction, a collection of papers on the topic. And David Foster’s book on Generative Deep Learning becomes available for free.