Big Tech & Startups
Stories about tech giants, startups, and venture capital
Wikipedia's Jimmy Wales Launches Social Network (1 minute read)
A new social media network called WT:Social has been stealthily launched by Wikipedia co-founder Jimmy Wales. WT:Social was previously a news-focused site, but it was revamped and has slowly been onboarding new users since last month. Its focus is on removing bad actors, editing misleading stories, and giving people the choice to dictate which stories get presented to them. WT:Social will allow users to build niche communities called SubWikis that will rely on volunteers to moderate and enforce standards of content. The platform survives on donations and is ad-free. There is currently a waitlist to join WT:Social. Users can skip the queue by pledging $12.99 a month or $100 for a year. None of the data generated on the site will be sold.
Google is scaling back its weekly all-hands meetings after leaks, Sundar Pichai tells staff (4 minute read)
Google will stop holding its weekly town halls, moving its TGIF meetings to just once a month. Low attendance and the lack of focus on product launches and business strategies were two reasons stated for the change. Employees were sharing the conversations held inside TGIF meetings outside of the company, which was affecting the ability of the company to hold candid conversations on important topics. TGIF meetings were previously a place where employees could bring concerns to management and discuss important workplace issues. Google will continue to hold regular Social TGIFs in offices around the world, and other forums and platforms have been provided for employees to communicate with the company. Employees can use these forums to voice their opinions on company decisions.
Science & Cutting Edge Technology
Stories about scientific breakthroughs and futuristic
technologies like AI, blockchain, and space travel
Novel AI System Can Predict Seizures with 99.6% Accuracy Up to an Hour Before They Occur (1 minute read)
An artificial intelligence system has been able to detect epileptic seizures in patients with 99.6 percent accuracy up to an hour before they occur. The system combines EEG technology and predictive modeling. It needs to be trained on each individual patient. Researchers are now working on customized hardware in order to make the technology more practical and comfortable for patients.
Boeing abandons its failed fuselage robots on the 777X, handing the job back to machinists (4 minute read)
Boeing spent six years on a project that implemented a large-scale robotic system for automated assembly of its 777 fuselages. The project cost millions of dollars and will now be abandoned in favor of human machinists. New techniques make the robotic approach obsolete. When it was introduced, the robot system was difficult to set up and produced fuselages that had errors and needed to be finished by hand. These issues have not completely gone away, and new techniques make using human workers a better option.
Programming, Design & Data Science
Tools, open source libraries, and other resources
for programmers, designers, and data scientsts
Hacking Neural Networks: A Short Introduction (GitHub Repo)
This repository contains short introductions on methods of finding vulnerabilities in neural networks and exploiting them. There are nine exercises covering topics such as backdooring, neural overflow, malware injection, and GPU attacks.
Package validator (GitHub Repo)
Package validator implements value validations for structs and individual fields based on tags. It features Cross Field and Cross Struct validations by using validation tags or custom validators, the ability to dive into both map keys and values for validation, alias validation tags, customizable i18n aware error messages, and more. Full documentation, benchmark scores, and examples are available.
Random stuff techies might be interested in
How I Created A $60K/Month App That Collects In-Person Payments Through Stripe (8 minute read)
Payment.co is an app built on top of the Stripe payment gateway to allow people to easily collect payments in person without barriers or card readers. It started in 2015 and currently processes $70 million in volume annually, taking in a 1% service fee on every charge. Ryan Scherf, the founder of payment.co, originally went to school to become a developer but transitioned later into design. In 2013, Scherf was approached by someone who had an idea for an app that was built on top of Stripe, and they launched Paid in 2013. After some success, Payment was launched. Scherf and his partner spent little on marketing or launching their product. They have since grown the company by building features based on customer feedback. Payment.co has no support team - Scherf does all of the support. There is still only one true employee in the company, but Scherf hires several contractors. Scherf's advice for other entrepreneurs is that there is no perfect time to start a company and that entrepreneurs should learn how to build products themselves, rather than saving money to get someone else to build it.