DataDownload: The “dark magic of the internet” that brought Facebook down
DataDownload: The “dark magic of the internet” that brought Facebook down A weekly summary of all things Media, Data, Emerging Tech View this email in your browser
Last week we gathered — and happily lots of you attended! For those of you who missed the Summit, we’ll have videos to share shortly — and… next year… we’ll be BACK in person (I make that promise, with my fingers crossed). Now, back to our regally scheduled newsletter.
First — Facebook’s week was going to be bad. When the WSJ, 60 Minutes, and Congress all have at you — that can’t be good. But then — kaboom, Facebook, Instagram, and WhatsApp all went down hard. And twice! Gizmodo has a solid take on what happened the first time, and it’s worth a read.
Speaking of sad tech — “Hey Siri, what time is it?” To which it responds… “I’m sorry, I didn’t get that.” Siri’s fall from the limelight is gradual, but unmistakable. The Verge digs in to why Siri isn’t what it promised to be.
Dark Design, as Fast Company reports — is flat out evil. If I have to find more crosswalks or traffic lights to prove I’m a human, I’m going to reach into the internets and strangle a website designer. Really.
And if you haven’t seen the Atlas robot by Boston Dynamics — well, here it is. We’ve watched this a bunch of times — and it’s like watching Sci-Fi come to life — if only they’d program it to say… “I’ll be back.”
WNYC has a podcast about The Hidden Biases of Search Engine Algorithms that’s worth a listen. And finally, if you want to know the job that is most likely to be taken by AI, look no further than the MIT Tech Review piece on automated Weathercasting. Now just hook that up to Epic’s MetaHuman and Al Roker better watch out!
That’s it — feeling optimistic, not sure why. Send us your feedback, thoughts or ideas. We’re always listening.
The NYC Media Lab Must-Read ‘The Dark Magic of the Internet’ That Brought Facebook to Its Knees
So… what the hell happened? Feels more appropriate to lead with that. For five hours on October 4th, Facebook’s network was completely disconnected from the internet, pulling WhatsApp and Instagram with it. Conspiracy theories abound — but eventually the company did release a statement, and Computerphile had some fun exploring the snafu (see video below).
In a nutshell, Facebook’s day off was kicked off with a faulty configuration change that accidentally “shut down Facebook’s backbone, the globally distributed network of fiber optic cables responsible for connecting all of the company’s data centers throughout the world.” Of course, it’s more complicated than that, having to do with the Border Gateway Protocol, the “glue that holds the web together…. fully understood by no one.” We recommend both Gizmodo’s explanation and the video for this one.
Gizmodo / 6 min read
A decade ago, Apple Fellow Phil Schiller stood on stage and announced that the interface of the future was here. For decades, engineers and researchers had been trying to create tech we could actually talk to, said Schiller, but the reality always fell short. Then, after a pause, he went on to the next slide, where he introduced something called Siri. The iPhone 4s voice assistant was an instant hit during a difficult time for Apple (Steve Jobs passed away a day after Schiller’s announcement). But Siri’s star power waned over the years, and now the assistant is overshadowed by Google Assistant and Alexa. How did Apple miss its shot?
The faults highlighted by Ars Technica in 2011 remain true to this day, but novelty and wow factor made them forgivable ten years ago. “Looking through reviews and comparisons of digital assistants in this period, two things stick out. The first is that people soon get bored of Siri…. The second major trend is that once competitors did arrive, Apple’s advantage evaporated quickly…. Looking back, it’s clear that Siri’s big problem is that it failed to maintain momentum.”
The Verge / 13 min read Read More Tech+Media Balenciaga Takes Springfield
Isn’t it ironic that it took a Balenciaga-sponsored episode to bring The Simpsons, however briefly, back to form? Actually, in all honesty, we haven’t really kept tabs since the 2000s, but this was still a surprise. The 10-minute episode was a thematic part of creative director Demna Gvasalia’s role reversal at this month’s Balenciaga show, “in which the fashion audience, usually in the position of judge and curator, was pretty brilliantly swapped into the role of star and then, just as quickly, lulled back into the thankless role of spectator…. Some of these models were actual celebrities, while others, including many Balenciaga employees, were just treated like them, which made it pointless to tell the difference.”
We’re familiar with Springfield’s brand of biting satire, but it was nice to see a fashion label poking fun at itself — “a rare thing in these self-serious fashion times (or actually ever),” as GQ notes. (Also see: Balenciaga x Fortnite hoodie.)
GQ / 8 min read Read More Amazon’s Astro Robot: A Feat of Science or a Successful Product? VentureBeat is calling Astro, Amazon’s latest robot assistant, an “Echo Show on wheels” (and Colbert calls it a dog’s worst enemy). But the Wall-E-like assistant is an amalgam of years of innovation in AI and sensor tech that, eventually, might combine into something beyond our current-gen voice companions. Feature-wise, Astro does what you’d expect of current-gen smart home tech (just, on wheels): “[It can] play music and video, and provide remote monitoring of different parts of your home. It can recognize your home’s residents and detect strangers. It can react to unusual events such as the sound of breaking glass.” Mobility is another interesting aspect of Astro:
“The robot must detect walls, doors, objects, people, animals, obstacles, and stairs, and create a map of your home to navigate it. In robotics, this is called simultaneous localization and mapping (SLAM), and it’s a longstanding challenge of the field. Astro solves SLAM by using multiple cameras and sensors to map its surroundings and runs them through several neural networks and planner systems. Navigation, in particular, needs to happen on the device, because a roundtrip to the cloud would be expensive and too slow to deal with the real-time changes that can happen in a house.”
VentureBeat / 8 min read Read More How to Spot the Psychological Manipulation Behind ‘Dark Design’ Online
We’ve all stumbled on the notoriously benign cookie banner, giving us two choices: click on the nice accept button to make it go away, or go through a maze of small-text options to safeguard your privacy. Frustrated, most of us choose the former. This is just a small example of the web’s plethora of dark designs — “the practice of creating user interfaces that are intentionally designed to trick or deceive the user.” Another example is airline sites that insidiously check off extra options by default, inflating your final ticket bill (and making it annoying to go back); then there are more subtle manipulations, like placing ads amidst videos on YouTube or posts on Reddit.
Fast Company / 5 min read
“Parkour is the perfect sandbox for the Atlas team at Boston Dynamics to experiment with new behaviors. In this video our humanoid robots demonstrate their whole-body athletics, maintaining its balance through a variety of rapidly changing, high-energy activities. Through jumps, balance beams, and vaults, we demonstrate how we push Atlas to its limits to discover the next generation of mobility, perception, and athletic intelligence.”
Boston Dynamics (YouTube) / 1 min watch
Watch Now What We’re Listening To Podcast: The Hidden Biases of Search Engine Algorithms
“Whenever a user enters a term into a search engine, they’re likely to be flooded by millions of results in a split second. Having so much content available at our fingertips can feel thrilling. But what’s less apparent is how much bias is baked into the algorithms that bring certain results to the top of our Google searches, while suppressing others.”
WNYC Studios / 12 min listen
Listen Now Virtual Events Free Event: Verizon 5G Edge Lightning Talks
Date: October 13, 2PM EDT
“We’re hosting our second ever 5G Edge Lightning Talks: an opportunity to bring together 5G Edge practitioners to share best practices, lessons learned, and practical advice for developers hoping to get started!” Register Here. A Deeper Look DeepMind’s AI Predicts Almost Exactly When and Where It’s Going to Rain
For a while, DeepMind’s milestones weren’t much different from OpenAI’s — occasional, stunning advancements in AI — some even the subject of excellent documentaries — that were not practical outside of a very narrow domain, or were toy examples. But AlphaFold 2 — DeepMind’s protein-folding prediction system — broke the mold. Arguably, it was deep learning’s biggest real-world impact since the 2012 ImageNet moment. And it’s what CEO Demis Hassabis was aiming at for decades. DeepMind continues to tackle hard science problems, including something that’s always been particularly tricky for meteorologists: predicting the likelihood of rain in the next 90 minutes.
Outdoor events, filming, aviation, emergency services, and other industries rely on weather forecasts. But weather is hard: “Figuring out how much water is in the sky, and when and where it’s going to fall, depends on a number of weather processes, such as changes in temperature, cloud formation, and wind. All these factors are complex enough by themselves, but they’re even more complex when taken together.” DeepMind developed DGMR, a deep learning tool that is able to predict the location, extent, movement, and intensity of rain 90% of the time. In a blind comparison with 56 forecasters, 89% preferred DGMR’s results over both a state-of-the-art physics simulation and a rival deep-learning tool.
MIT Technology Review / 4 min read