This edition has 344 words which will take you about 2 minutes to read.
“But nothing is more opaque than absolute transparency.” – Margaret Atwood.
Data Exchange podcast
- Towards Simple, Interpretable, and Trustworthy AI I speak with Sheldon Fernandez, CEO at Darwin AI, and Alex Wong, Professor at the University of Waterloo, Co-Founder of DarwinAI and Euclid Labs. They recently published a series of research papers and articles introducing new tools for improving trust and transparency in deep learning.
- The Rise of Metadata Management Systems Assaf Araki, investment manager at Intel Capital, explains why metadata will be the foundation for data governance solutions, data catalogs, data lineage, data quality, and other enterprise data applications.
Free Report
These are insightful conversations about best practices for building Data and AI products & services. This compilation features technical leaders from Shopify, Rakuten, Bloomberg, and Fiddler.
Data & Machine Learning tools and infrastructure
- A summary of data processing support in Ray Ray provides developers a single substrate for distributed data processing and machine learning.
- GraphScope from Alibaba A unified distributed graph computing platform, that makes graph operations and multi-staged processing of large-scale graph data easy and accessible through a Python interface.
- Monitoring and MLOps in the context of Facebook’s Newsfeed
- A preview of Julia 1.6 For more on Julia, listen to this Data Exchange podcast episode with Julia co-creator Viral Shah.
- DeepMind’s new state-of-the-art image recognition models, trained without batch normalization, are also 8.7x faster to train
Funding Updates
- TigerGraph raises $105 million in Series C funding
- Matillion raises $100M Series D Round Data integration is a hot area, there are many well-funded startups working on ELT, data pipelines, and DataOps.
Recommendations
- One Simple Chart: online learning platforms, a year into the pandemic
- Patterns, Predictions, Actions A fantastic new graduate textbook in machine learning by Moritz Hardt and Ben Recht.
- Don’t Get Burned Shorting Bubbles There are bubbles everywhere ⎼ BTC, TSLA, to name a couple. But actually putting on trades to profit from bubbles requires skill, patience, and fortitude. This paper explains why many investors passed on the famous “short subprime” trade from 2005 to 2008.
- This Is How They Tell Me the World Ends: The Cyberweapons Arms Race
If you enjoyed this newsletter please support our work by encouraging your friends and colleagues to subscribe: