11/21/17: Net Neutrality, Chinese Robot Surgeons, Understanding Neural Networks.

Time for the Daily Decode! Here are three links worth your time.1.

Via The New York Times: F.C.C.

Announces Plan to Repeal Net Neutrality by Cecilia KangTL;DR:Basically, Internet providers will act like this.Why is this important:Remember, ISPs have already successfully lobbied to be able to sell your private internet data, aka where you have been, to the highest bidder. Since, they cant compete with companies like Google or Facebook who actually have to have a product to get that extra data, they decided they will just pass a law to get it automatically without any additional innovation.

Also remember that ISPs have record profits already and the highest paid CEO in America is from Charter Communications. His take home salary is $98 million dollars, up almost 500% from last year. And to top it off, ISPs have passed laws to prevent municipalities from creating their own lower cost ISPs, or set so many barriers to slow down the progress and crush competition.

That means less competition and a little choice in most areas. Furthermore, our internet lags behind most countries with the US paying more for less services. Telecoms and the FCC chairman say that net neutrality rules have prevented it from expanding in areas, thus limiting investment.

Yet, the data doesnt point directly to that. Some say the chairman is even ignoring the actual data. Fundamentally, the power of the internet is that its an open and permission-less platform where anyone can compete.

Erecting a gatekeeper and trusting them to not act in a selfish way to stymie the competition is foolish, damaging to our economy long term at worst. Do you honestly think the ISPs wont try to recoup the money lost as people cut their cords to cable service and sign up for Netflix? Overall, this is pointing towards lowering the innovation of the internet economy for the sake of boosting the profits of a few gate keepers.

So. now this. Is that enough for you to call your local Congressman and Senator?

2. Via Futurism: For the First Time, a Robot Passed a Medical Licensing Exam by Dom GaleonChinese AI-powered robot Xiaoyi took the countrys medical licensing examinations and passed, according to local reports. Xiaoyi is just one example of how much China is keen on using AI to make a number of industries more efficient.

Do you think the rent seeking behavior in America will allow this? Send you responses to ai.at.

or start a discussion in the forums.3. Via omar-florez.

github.io: One LEGO at a time: Explaining the Math of How Neural Networks Learn By Omar U. FlorezWhy should you read this:If you understand the internal parts of a neural network, you will quickly know what to change first when things dont work and define an strategy to test invariants and expected behaviors that you know are part the algorithm.

This will also be helpful when you want to create new capabilities that are not currently implemented in the ML library you are using.Events:Free Python Open Workshop:Sunday, November 26th, 2017Fordham University School of Law113 West 60th StreetRoom 519New York, NY 10023We will continue our coverage of lists, move on to dictionaries, work on hacker rank problems, and introduce object oriented programming again. For those more advanced we will continue to work on Python Poker app and notes on how to use python with databases will be shared.

More advanced courses:Rahul Remanans will also be hosting a Natural Language Processing Course:January 6th, 201810am to 6pmLocation: TBDPrice: TBDThis 6 to 8 hour event will cover the best practices and an in-depth analysis of natural language processing using deep-learning. The following topics included are:1) One hot encoding, 2) Recurrent neural networks using long short-term memory (LSTM), 3) Natural language generation,4) Character embedding,5) Natural language classification,6) Conversational chat-bots using deep-learning.7) Understanding healthcare data using NLP.

That is all!Thanks,Anthony AlbertorioOrganizer w/ AI@CUMC