Big Tech &
Uber sues California over the gig-worker law. Now what? (4 minute read)
Ride-hailing and delivery drivers from Uber and Postmates are suing the state of California over a law that is set to take effect today. AB 5 introduces a test for determining when a worker must be classified as an employee and therefore be entitled to minimum wage and benefits. This law has severe implications for the business models of companies that rely on contractors. The lawsuit is an attempt to get lawmakers to rethink the law to create a better solution. Some workers want to remain classified as independent contractors due to the schedule flexibility it provides them. Experts believe that it is likely the lawsuit will be unsuccessful. Workers from many industries are affected by this law, including truck drivers, freelance journalists, and musicians. Some of these groups have also filed lawsuits.
Thousands of Google’s cafeteria workers have unionized (6 minute read)
Around 2,300 Google cafeteria workers have unionized, claiming that they are overworked and underpaid. The unionization is some of the most significant union activity that tech industry workers have accomplished. Cafeteria workers earn wages that start around $35,000 and don't receive the same benefits that are standard for full-time Google employees. On-site food workers are contracted to work via a third party, Compass Group. A local chapter of Unite Here is helping the workers organize their union. Service workers in the tech industry have seen stagnating wages that haven't kept up with the high cost of living. Some employees live in an RV camp that has formed outside Google's headquarters. Google announced in June that it would invest $1 billion in land and building homes.
The family in 2050: artificial wombs, robot carers and the rise of single fathers by choice (10 minute read)
Many of the predictions about what 2020 would be like for families didn't pan out, but things have definitely changed in the last couple of decades. Women are working more, the number of people choosing not to have children is increasing, and same-sex and transsexual parents are becoming more common. As reproductive technology develops, the definition of a typical family is likely to change significantly. Science will soon be able to create artificial sperm and eggs, as well as wombs. A change in the process of reproduction may mean that women can continue to work while being 'pregnant', and people who could not produce their own children previously (eg. single men, transgender and same-sex couples) can have their own children. Genetically modified designer babies will also likely become a reality. Many other factors such as an aging population, increased travel, and economic inequality will also affect the future of the family unit.
When your next door neighbor is a glittering spaceport (4 minute read)
Spaceport America is a $220 million facility based in the deserts of New Mexico. If everything goes to plan, it may launch its first spaceships in 2020. Spaceport America's primary tenant is Virgin Galactic. Virgin Galactic sells seats for its spaceplane at $250,000 a ticket, and around 600 people have paid a deposit so far. Passengers will spend a few days at the spaceport to train before their trip. A nearby town approved a tax in the last decade to help build the spaceport as it promises to bring an influx of business to the area. Virgin Galactic has taken longer to get into space than originally expected, with a fatal in-flight accident in 2014 causing more delays to the program. It has since flown its spaceplane to the edge of space and back twice.
Programming, Design & Data Science
Neural Network Intelligence (GitHub Repo)
Neural Network Intelligence (NNI) is a toolkit to help users automate Feature Engineering, Neural Architecture Search, Hyperparameter Tuning, and Model Compression. It was designed so that researchers and data scientists could try different AutoML algorithms and environments for their models. NNI comes with a command-line tool and a WebUI.
Ask HN: Full-on machine learning for 2020, what are the best resources? (Hacker News Thread)
There are many topics to choose from when starting out in machine learning. It can get confusing, and the theory can be quite difficult to learn. Many people suggest that learning using a 'top-down' approach might be best. People can learn how models work and become productive much faster by just diving into the code. Theory can be learned concurrently, or even at your own pace after you've learned the practical parts. Learning theory can help you build higher quality models. There are lists of recommended resources for machine learning in the thread.
How the Entertainment Industry Solved Piracy, Then Made It Popular Again (6 minute read)
For the last few decades, the entertainment industry has been fighting piracy by suing and vilifying potential customers or lobbying for copyright laws like DMCA, rather than offering better, cheaper products. The introduction of streaming services saw a drop in piracy as people had an attractive, legal alternative. As more and more streaming services enter the market, consumers are forced to sort through the many different offerings to find the shows that they want to watch. If consumers want to watch shows that are on different platforms, they have to pay for both. This is causing some consumers to return to piracy. Data suggests that piracy is an invisible competitor and a metric of consumer dissatisfaction.
No TLDR Originals for 2020-01-01