This edition has 572 words which will take you about 3 minutes to read.
Recommendations:
Inspired by Lauren Greenfield’s excellent documentary (“The Kingmaker”), I collected some quotes on disinformation that might prove handy in this election year in the US. And if you’re looking for a great documentary: ”The Kingmaker” was one of my favorite films from 2019.
Sandworm is an excellent new book by Andy Greenberg of Wired Magazine. It describes chilling attacks on computer systems and industrial control systems, the likes of which we might see more of in 2020 (particularly in light of recent events in the Middle East).
Mikio Braun and I go out on a limb and speculate about trends in AI and Data that we think people should pay attention to in 2020.
Episode 4: Key AI and Data Trends for 2020
Subscribe: iTunes, Android, Spotify, Stitcher, Google, and RSS.
Bahman Bahmani is VP of Data Science and Engineering at Rakuten, a large Japanese ecommerce and online retail company. In my opinion he has established an organizational structure, processes and a data science practice that other companies should study.
Software + Commodity Hardware should be able to handle deep learning
In a conversation that took place at AI Week in Tel Aviv, Nir Shavit of MIT and Neural Magic explains why he thinks new algorithms and software will allow commodity CPUs to achieve reasonable enough performance, lessening the need for specialized hardware for deep learning. While this is currently somewhat of a contrarian position, I’m looking forward to reviews of Neural Magic’s initial product (for inference in computer vision).
Machine Learning tools and infrastructure
A couple of technology companies just released code and articles about their internal ML tools and platforms. Within Linkedin, Pro-ML is a set of initiatives intended to “double the effectiveness of machine learning engineers while simultaneously opening the tools for AI and modeling to engineers from across the LinkedIn stack.” Meanwhile, Uber open sourced Manifold, an interesting visual debugging tool for machine learning.
Ray RLlib stacks up quite well in this comprehensive comparison of open source tools for reinforcement learning.
Conferences :: Call For Speakers
A couple of SF Bay Area conferences I’m involved with are still looking for speakers:
- Ray Summit (“Scalable machine learning, Scalable Python, for everyone”): deadline to submit is January 31st. This is a new conference centered on Ray an open-source system for scaling Python applications from single machines to large clusters. Based on what I know about the program, this is going to be an outstanding inaugural event.
- Apache Pulsar Summit: deadline to submit is January 31st. For more on Pulsar, see this Hacker News thread from January 2nd.
If you are going to be in the San Francisco Bay Area next week, there will be a Ray Meetup on January 30th in San Francisco. Topics include Ray.serve (scalable machine learning model serving on Ray) and reinforcement learning. See you there!
Work and hiring:
- The tech talent pipeline: a recent study from McKinsey compares key metropolitan areas in the US.
- Here’s an interesting thread on work/life balance from Shopify’s CEO.
- Hiring for data roles: An in house recruiter at GetYourGuide describes his approach to recruiting data professionals in an environment when many companies are going after the same talent.
Subscribe to our Newsletter:
We also publish a popular newsletter where we share highlights from recent episodes, trends in AI / machine learning / data, and a collection of recommendations.
[Image: Newsletter from Pixabay.]