Profile cover photo
Profile photo
Wayne Radinsky
Software Design Engineer
Software Design Engineer

Wayne's posts

Post has attachment
A neural network to distinguish voices made by humans from voices made by neural networks. It's basically a neural network trained on voices made by humans and neural networks. It works for now, but as neural network voices get more and more similar to real human voices, I question whether it is going to be able to continue to distinguish the two, just as it will get more and more difficult for our human brains to distinguish the two.

Post has attachment
"Apple is working on a dedicated chip to power AI on devices." "The chip, known internally as the Apple Neural Engine, would improve the way the company's devices handle tasks that would otherwise require human intelligence -- such as facial recognition and speech recognition," according to rumor.

"Should Apple bring the chip out of testing and development, it would follow other semiconductor makers that have already introduced dedicated AI chips. Qualcomm Inc.'s latest Snapdragon chip for smartphones has a module for handling artificial intelligence tasks, while Google announced its first chip, called the Tensor Processing Unit (TPU), in 2016. That chip worked in Google's data centers to power search results and image-recognition. At its I/O conference this year, Google announced a new version that will be available to clients of its cloud business. Nvidia Corp. also sells a similar chip to cloud customers."

Post has attachment
The top 5 companies in the world by market cap are all tech companies. I don't think I've ever seen this before. In fact I'm used to seeing an oil company -- usually EM -- at the top of the list, and tech companies scattered about the top 10. But right now, the top 5 companies in the world are (in this order):

Alphabet (Google)

Have we entered a new era, folks?

Post has attachment
"At Facebook, every change made to our mobile code is checked by our static analyzer Infer. Static code analyzers are tools that scan the source code of a program to detect possible bugs. The main advantage of a static analyzer is its ability to detect bugs without running the program and before the software is shipped to the users."

"One benefit of Infer over other static analyzers is that it performs sophisticated interprocedural/interfile analysis at scale. That is, Infer can detect subtle bugs by tracking values through many procedure calls or where the procedures live across different files. Remarkably, although sophisticated and deep, Infer's analyses scale well to large programs consisting of millions of lines of code."

"Despite Infer's advantages, one of its limitations has been its extensibility. Adding a checker for a new type of bug was a complex task and required a lot of deep static analysis expertise as well as knowledge about Infer's internals. This was especially painful for writing new checkers that didn't require all the sophistication that Infer offered. In fact, many significant software bugs are confined to the code of a single procedure (called intraprocedural). To detect these bugs, it may suffice to perform simpler analyses that do not require technical expertise in static analysis. These bugs can often be expressed by referring to the syntax of the program."

"For this reason, we have introduced a new language called AL to easily design new checkers for detecting these kind of bugs."

Post has attachment
Is China Outsmarting America in AI? "National efforts are aided by access to enormous amounts of data held by Chinese companies and universities, the large number of Chinese engineers being trained on either side of the Pacific and from government backing."

"Driving that attention is a breakthrough from an American company largely banned in China: Google. In March 2016, a Google artificial intelligence system, AlphaGo, beat a South Korean player at the complicated strategy game Go, which originated in China. This past week, AlphaGo beat the best player in the world, a Chinese national, at a tournament in Wuzhen, China. The Google event changed the tenor of government discussions about funding."

Post has attachment
50 AlphaGo vs AlphaGo games for Go players to enjoy.

Post has attachment
Amazon "launched Amazon Charts, which complements the site's usual hourly updates of bestselling books. The new list combines what's being ordered from them with data obtained from Kindle and Audible users to find out what books are actually being read and listened to."

"It's an interesting algorithm, and one that has been utilised before, but never formally by Amazon in this way. In 2014, the mathematician Jordan Ellenberg created an index of the most abandoned books, based on Kindle data. So while every man and his dog might have bought a copy of Stephen Hawking's A Brief History of Time and Thomas Pinketty's Capital in the Twenty-First Century, not everyone actually read them."

Post has attachment
"There's a limit to how fast autonomous vehicles can fly while safely avoiding obstacles. That's because the cameras used on today's drones can only process images so fast, frame by individual frame. Beyond roughly 30 miles per hour, a drone is likely to crash simply because its cameras can't keep up."

"Recently, researchers in Zurich invented a new type of camera, known as the Dynamic Vision Sensor (DVS), that continuously visualizes a scene in terms of changes in brightness, at extremely short, microsecond intervals. But this deluge of data can overwhelm a system, making it difficult for a drone to distinguish an oncoming obstacle through the noise."

"Now engineers at MIT have come up with an algorithm to tune a DVS camera to detect only specific changes in brightness that matter for a particular system, vastly simplifying a scene to its most essential visual elements." The results "can be applied to any linear system that directs a robot to move from point A to point B as a response to high-speed visual data."

Post has attachment
11 chatbot companies to watch from Y Combinator's Startup School. (enterprise-level bot building platform), chatShopper (fashion chatbot on Facebook Messenger), BotMakers (Chatbot Marketplace that connects businesses with bot developers), botanalytics (conversational analytics for bots to help improve human-to-bot communication), Weps (AI to build your website instantly), BusyBot (manages tasks in Slack), Tars (chat like interface for filling out forms on mobile), Botsociety (tool to design chatbot appearance for marketing materials, promoted Facebook posts, etc), Hunch (analyzes complex acquisition, retention, referral, and revenue data and reports in easy to understand language), Rocco( AI social media marketing assistant on Slack), and Botsify (creates rule-based chatbots for Facebook Messenger and Facebook business pages).

Post has attachment
Google made a "digital whiteboard" called Jamboard that shares data through "the cloud." I don't know how it detects writing on the board. It looks like it might be using AI to translate writing and doodles into text and diagrams.
Wait while more posts are being loaded