Artificiality Bites 💊 Issue #46

Hello Human! This is a new issue from my weekly newsletter, holding a tiny compilation made of interesting articles from last week, projects, tutorials and tools; all related to Data, Artificial Intelligence and adjacent topics. Uživajte u jelu!

📝 Interesting publications this week

• Applied NLP Thinking: How to Translate Problems into Solutions 19'
  This blog entry explains some of the biggest challenges for applied NLP and how to translate business problems into machine learning solutions.
• How to Do Multi-Task Learning Intelligently 11'
  This article discuss the motivation for MTL as well as some use cases, difficulties, and recent algorithmic advances.
• On the “usefulness” of the Netflix Prize 7'
In 2006 Netflix launched a $1M competition to improve their recommendation algorithm. This article provides insider details around the prize and its outcome.

💡 Projects

• GitHub Copilot
  An AI pair programmer powered by Codex, a new system created by OpenAI.
• Alias-Free GAN
  This new Generative Adversarial Network from Nvidia solves StyleGAN2’s aliasing problems, with renditions fully equivariant to translation and rotation even at sub-pixel scales, being more suitable for video and animation.

• NVIDIA Canvas
  Use AI to turn simple brushstrokes into realistic landscape images. Application now in Beta.
• NL-Augmenter
  A Collaborative Repository of Natural Language Transformations.
• LyricJam
  A system for generating lyrics for live instrumental music.
• Translated Wikipedia Biographies
  A dataset designed to analyze common gender errors in machine translation like incorrect gender choices in pro-drop, possessives and gender agreement. Read more at Google AI Blog.

🔧 Tutorials

• NLP from Scratch with PyTorch, fastai, and HuggingFace 35'
  A technical NLP tutorial using a variety of libraries to show the different levels/layers of common NLP pipelines.
• Sign Language Detection using Action Recognition with Python and LSTM 📹 2h27'
  How to build a Sign Language model using Action Detection powered by LSTM layers written in Tensorflow/Keras, and then make predictions in real time by using video sequences.
• A from-scratch tour of Bitcoin in Python 65'
  Create, digitally sign, and broadcast a Bitcoin transaction in pure Python, from scratch, and with zero dependencies.

📦 Repositories

• facebookresearch/kats
  Kats is a lightweight, easy-to-use, and generalizable framework to perform time series analysis, open-sourced by Facebook’s Infrastructure Data Science team.
• OpenLineage/OpenLineage
  OpenLineage is an open standard for metadata and lineage collection.
• nidhaloff/igel
  A machine learning framework that allows you to easily train, test and use models without writing code.
• microsoft/PythonProgrammingPuzzles
  A dataset of python programming challenges which can be used to teach and evaluate an AI’s programming proficiency.
• allenai/S2AND
  This repository provides access to the S2AND dataset and reference model for Author Disambiguation System in Semantic Scholar. More details in here.

🎓 Courses / Events / Books

• Introduction to Machine Learning Interviews Book 📕
  This book draft by Chip Hunyen is the result of the collective wisdom of many people who have sat on both sides of the table and who have spent a lot of time thinking about the hiring process.

🚀 Extra bits

• From admissions to teaching to grading, AI is infiltrating higher education 13'
• Amazon is using algorithms with little human intervention to fire Flex workers 6'
• After Repeatedly Promising Not to, Facebook Keeps Recommending Political Groups to its Users 10'
• Linux Foundation Launches Open Voice Network 4'
• Tinkoff Makes Voice Assistant Oleg Introduces Universal AI Phone Defense Against Spam Calls 3'
• Toyota announces new capabilities for domestic robots 3'
• Google Launches a New Medical App outside the US 8'
• Voice AI is scary good now. Video game actors hate it 7'

@TheInsaneApp

👋 See you next week!