Big Tech &
SpaceX will now let you book a rocket launch online starting at $1 million (2 minute read)
Customers who want to reserve space on a Falcon 9 launch can now do so through SpaceX's new web-based booking tool. SpaceX is offering rideshare services for $1 million per payload ranging up to 200kg, with an additional $5 per extra kilogram of cargo. Customers can select the desired orbit and the launch date, input the total amount and size of cargo, and select a specific rocket to book a ride on. The website will return an estimated cost. Customers can then choose some extra options, such as insurance. A small deposit is required, after which the customers can pay the rest of the fee in up to three installments. A Rideshare User's Guide has been released by SpaceX that contains more details about the program. Screenshots of the booking tool and a link to the Rideshare User's guide are available in the article.
Spotify is buying Bill Simmons’s The Ringer to boost its podcast business (4 minute read)
Spotify has bought The Ringer, a podcast-centric media company owned by Bill Simmons. Simmons and his employees will all be absorbed in the deal. The Ringer has around 30 titles, ranging from an NBA chat show to a show devoted to rewatching old movies. Neither company disclosed a sale price for the deal, which is expected to close in the first quarter of this year. Spotify did not make Gimlet Media's podcasts exclusive to its platform when it purchased the company last year, so The Ringer's podcasts should still be available on all platforms. This is the fourth podcast company Spotify has acquired in the last 12 months.
Researchers Say ‘Anti-Solar Panels’ Could Generate Power at Night (2 minute read)
Researchers from the University of California have designed 'anti-solar panels' which would theoretically be able to generate solar power at night. Solar panels work because they are cold compared to the sun, so they absorb sunlight and convert it into energy. The anti-solar panels reverse the process and generate energy through the loss of energy from the panels to the cool night sky. It is estimated that the panels would be able to generate about a quarter of the power of a solar panel. The next step for the scientists is to build the devices to see how well they perform. It may be possible to boost the efficiency of the cells by using materials that can better interact with longer wavelengths of light.
YouTuber uses neural networks to upscale 1896 short film to 4K 60 fps (3 minute read)
A YouTuber was able to use a neural network to upscale a very old, low-resolution, black and white video from 1896. The result was a clear 4K video that played at 60 frames per second. AI has been a hot topic in recent years, with the technology being applied to create technology such as autonomous cars. Denis Shiryaev used a mix of neural networks from Gigapixel AI and a technique called depth-aware video frame interpolation to improve the quality of the old video. The Gigapixel AI was trained to 'fill in' information in an image to upscale it and improve its quality. It is possible that the AI could introduce erroneous details, and it is much slower than conventional techniques. The network can be trained to colorize images and add in various missing patterns in objects like floors, windows, and vegetation. Some data scientists are attempting to create neural networks that can draw and paint. AI is getting so big that large tech companies are all racing to buy or invest in every new and obscure startup in the field.
Programming, Design & Data Science
Scaling to 100k Users (8 minute read)
As startups grow, they will inevitably have to deal with issues of scale. This can be a good problem, but there is scarce information on how to scale web apps as the number of users grows to hundreds of thousands. Developments usually come from dealing with massive issues or identifying bottlenecks. As a web app grows, it will have different needs. Infrastructure is built and changed incrementally. Larger web apps will start needing services like load balancers and CDNs. As businesses continue to grow, developers will always need to continue to work to solve problems of scale.
Experienced data scientists, what are the most common issues you see newly graduated students have when they start working for the first time? (Reddit Thread)
The most common issue for new data scientists is that they don't question their own results. It is important to validate the results that are generated from machines as the models aren't always correct and they may make the wrong predictions. A good data scientist is critical of their own work and will seek to find out where they might have gone wrong. Newer data scientists should learn how to estimate processing times, follow basic coding practices, learn how to say 'no' especially if the task is impossible, value product knowledge, learn the industry, and learn from others in the industry. If a data scientist is unfamiliar with the business they are working for, the results and predictions they generate may be useless for their company.
Extreme Silicon Valley: A 2:30 a.m. bus from Salida (8 minute read)
Tech companies are operating shuttle buses that go all the way to the Central Valley as more and more tech employees move into the area. When tech companies first started offering shuttle buses for their employees more than a decade ago, these buses were focused in affluent neighborhoods. Now they reach much further into city outskirts. Housing prices have risen sharply where tech companies are based. Anti-development politics in Silicon Valley, San Francisco, and the East Bay make it hard to build new housing and improve public transit systems. Some large tech firms, such as Google and Facebook, have proposed building apartment blocks on their own campuses.
Here Is a Link to the App that Blew Up the Iowa Caucus (7 minute read)
The Democratic party has delayed the announcement of the winner of the Iowa caucus due to a failed app that caused vote numbers to be reported incorrectly. Prior to the election, the party had kept the app developer's identity secret for security purposes. It has now been revealed that the app was developed by a startup called Shadow, which has also previously provided services to the Texas Democratic Party, Kirsten Gillibrand's primary bid, and Pete Buttigieg's campaign. Shadow's parent company, Acronym, has stated that they only invested in the company and that they don't know the specifics of the app failure. The app was never vetted by major cybersecurity agencies or the wider cybersecurity community. Many of the specifics of the app are still kept secret, which is a cause for concern according to security experts.