The old monthly [Blog Reads] is going to be replaced by TIL! Basically, these posts will have a list of interesting stories, blog posts and research that I found and want to talk about. Hence, I will not be limiting myself to one post per month. Furthermore, I will try to balance these TIL lists to be longer and more informative. Enjoy!
1. Google Deepmind’s Team had worked to create an AI that can self-learn to walk
The AI has virtual sensors which allows it to see the obstacles in front of it. It can then try different bone joining movements until it passes through the obstacle. The video itself is totally amazing to watch and makes us question just when are real-life walking robots going to become reality…
2. Network graph of Twitter Interactions between Left and Right Wings
You can read more about this trending topic on its official reddit post here. Since the 2016 Elections, I have seen and participated in countless debates on what went wrong and why nearly no one expected such results in favor of current U.S. President Donald Trump. One of my personal hypothesis (share with many) is that the lack of concrete and open communication between the American population caused a wrong representation of the popular opinions.
3. The data that transformed AI research—and possibly the world
Read this article to learn about IMAGENET, one of the most important image dataset to computer vision. Stanford professor Fei Fei Li played a crucial role in the exponential evolution of computer vision and deep learning for the last decade.
4. Disney turns to AI to track filmgoers’ true feelings about its films
Here is a new applied aspect of computer vision and human facial analyzes. I love to see the big players of various different fields get interested in the potential of AI for their product.
5. A beautiful data animation shows the unprecedented development of China’s rail system
It is no surprise that I love data visualization. Shown in aesthetic ways, data can make us understand so much more! This animation makes us realize just how fast China’s metro system is developing.
Less than 20 years ago, only three cities in China had subways. If you didn’t live in Beijing, Guangzhou, or Shanghai, underground travel was a complete mystery. Today, there are more than 60 metro lines(paywall) in 25 cities, making subterranean transit accessible to some 291 million people.
6. Auditing Search Engines for Differential Satisfaction
Are search engines fair? This question makes me wonder as Google is monopolizing the way we search, the information we see first. Whatever search engine you use, there is bound to be some bias when results are shown. This post talks about using various models to understand the best search result satisfaction.