Big Tech &
Yelp now offers personalized results based on your diet and lifestyle (2 minute read)
Users will be able to start selecting specific interests for restaurants and businesses in the Yelp app on iOS and Android. The feature will roll out over the course of the next year. Yelp will categorize businesses based on information from user-generated reviews, photos, questions, and answers. The Yelp app will provide personalized search results using a machine learning algorithm. This is the first time Yelp has offered personalization options, and it shows how the company is expanding its focus beyond food. Businesses won't have to do anything on their side.
Apple releases iOS 13.1 beta before iOS 13 is even out (2 minute read)
Apple has released the iOS 13.1 beta, weeks before its September 10th event where it is expected to announce the iPhone 11 to the world. The early beta release may be due to Apple relieving pressure on their iOS 13 promises by spreading out the development process over the entirety of iOS 13, as well as allowing the company more time to fix bugs in the system. A changelog is available that lists features such as AirPods Audio Sharing, which appeared in previous versions of the iOS 13 beta. This may mean that the feature may be released with iOS 13.1, rather than iOS 13.
Researchers create breathalyzer that can detect marijuana (3 minute read)
As marijuana laws change, the technology to enforce DUI laws has to evolve as well. A team at the University of Pittsburgh has developed a breathalyzer device that can measure THC in the user's breath. Current THC detection methods require more involved procedures that can't be performed infield, and can only detect whether the user has recently used the drug rather than if they were currently under the influence. The new breathalyzer binds THC molecules to nanotubes, which changes the electrical properties of the nanotubes. It is taught to recognize the changes in electrical current caused by THC through machine learning. The prototype of the device looks similar to an alcohol breathalyzer, and researchers hope to be able to make it available for use soon.
A Successful Artificial Memory Has Been Created (5 minute read)
Scientists have been able to create artificial memories in lab mice, resulting in mice that responded to a stimulus that they haven't been trained on before. In the experiment, mice were artificially trained to associate the smell of cherry blossoms with a foot shock. The mice had their cherry blossom smell-detecting nerves stimulated artificially along with nerves that were important for the aversive reaction to foot shocks. They were not exposed to the actual smell of cherry blossoms or real shocks. When tested, the mice responded to the real cherry blossom odor by showing avoidance. The study showed that the stimulation in these specific areas in the brain led to activity in other regions of the brain that were associated with memory formation. Memory transfer has been observed before in sea slugs through RNA and has been partially successful in experiments involving rodents through transferring recordings of electrical activity. However, in this study, the memories were completely artificial.
Programming, Design & Data Science
Gamedev Tutorial: Trigonometry Basics – Sine & Cosine (10 minute read)
Trigonometry is a large part of game math, and a solid understanding of it can go a long way in game development. It is essential for knowing how to define coordinates as points on a circle, which allows developers to program how things move in circles and spirals, as well as different types of motions. Many visual examples of movements and the math to create them are provided in this tutorial.
OpenSpiel: A Framework for Reinforcement Learning in Games (GitHub Repo)
OpenSpiel is a collection of environments and algorithms designed for research for general reinforcement learning and search/planning in games. It is able to play a wide range of games and game formats. While OpenSpiel has only been tested on Linux, there should be no issues running the software on Windows or Mac OS platforms. A list of compatible games is available, which includes Backgammon, Chess, and Tic-Tac-Toe.
Former Star Google and Uber Engineer Charged With Theft of Trade Secrets (7 minute read)
Anthony Levandowski, one of the pioneers in self-driving car technology, has been charged with 33 counts of theft and attempted theft of trade secrets from Google. He has posted a $2 million bail and is currently required to wear an ankle monitor as he is deemed a flight risk. Disputes between companies and former employees are common in Silicon Valley, but criminal charges against a senior executive for theft is unusual. It is alleged that Levandowski downloaded more than 14,000 files containing critical information about Google's autonomous vehicle research in 2016, including schematics for proprietary circuit boards and designs for Lidar. Levandowski is due in court on September 4, and if he is convicted, he faces up to 10 years in prison and a $250,000 fine for every count of theft, plus additional restitution. His lawyers maintain that he did not steal anything in over a decade of being an industry-leading innovator in the field of self-driving car and truck technology.
No TLDR Originals for 2019-08-28