Artificial Intelligence #40

Hey, in this issue: how machine learning improves Google search, first many to many machine translation system (with code), learning with practically no data, detecting deepfakes with 80% accuracy, image-captioning AI better than human, machine-learning-based fast COVID-19 test, and more.

This issue is sponsored by AMAX.

• Google details how it’s using machine learning to improve search (blog)
• Facebook released the first many-to-many multilingual model that translates between 100 languages without relying on English data (code, paper)
• AdaBelief optimizer: fast as Adam, generalizes as good as SGD, and sufficiently stable to train GANs (paper, code, video)
• Deconstructing Maxine, Nvidia’s AI-powered video-conferencing technology
• [Sponsored] AMAX - NVIDIA? DGX? A100 – One system for all AI workloads

Crunch through terabytes of datasets in seconds. Harness data insights in days. NVIDIA DGX A100 delivers unprecedented acceleration and flexibility to power AI, analytics, and HPC applications. Deployed by AMAX AI engineers for a seamless integration plus 24x7 support. Special pricing for academia/startups. Kick start your AI insights >

• A radical new technique lets AI learn with practically no data
• When do we trust AI’s recommendations more than people’s?
• A new neural network can detect deepfakes at about 80% accuracy (for now)
• Nemo: data discovery at Facebook
• Microsoft’s image-captioning AI is describing pictures better than a human
• New Oxford machine learning-based COVID-19 test can provide results in under 5 minutes
• [Announcement] Machine Learning Engineering, the book by Andriy Burkov is now released on Leanpub and Amazon (paperback, hardcover, and Kindle)

If you build an AI or data product or service, you are invited to sponsor one of the next issues. Feel free to contact True Positive Inc. for more details on sponsorships.

Enjoy the newsletter? Help us make it bigger and better by sharing with your colleagues and friends.

Have a nice weekend! See you next week. — Andriy