Monthly Archives: August 2017

A smart sensor to automate Apple households

This wireless motion sensor is part of Elgato’s lineup of HomeKit-compatible accessories and can be used to trigger “scenes” and rules based on your movement (or the lack thereof) around your house.


Like most HomeKit accessories, the Eve Motion is a snap to set up. It runs on two batteries and uses Bluetooth low energy (BLE) to communicate with your other iOS devices and Apple TV. No hubs or Wi-Fi setup required.

While you can use Apple’s Home app to configure the Motion, it makes more sense to download the dedicated Elgato Eve one, which gives you more options. After inserting the two included AA batteries into the sensor and launching the Eve app, you choose to add a new device and the Eve Motion should pop up immediately. After that, it’s simply a matter of scanning the unique HomeKit code on the sensor with your iPhone or iPad, selecting the room where you’ll be using it, and then updating the sensor’s firmware. All told, the process takes about six minutes.

You can also tweak the motion sensitivity and duration values during this initial app setup. The latter is used to determine how quickly the motion sensor will re-detect motion after it’s been triggered.

As with Apple’s Home app, you can use the Eve app to create “scenes” that execute multiple actions simultaneously with a single command (An “I’m Home” scene could trigger all the HomeKit-enabled lights in your home to turn on when the Eve sensor detects motion, for instance). You can also create timers and rules to set scenes automatically at specific times or with specific motion triggers.

To test the Eve Motion, we used it in various parts of our home—including the living room, basement, bedroom, and office—creating location-specific scenes, timers, and rules based on our movements.

Artificial intelligence now powers all of Facebook’s translation

Spend enough time on Facebook, and you’ll likely encounter a post written in a tongue that’s foreign to you. That’s because the social network has two billion users and supports over 45 languages. On Thursday, Facebook announced that all of its user translation services—those little magic tricks that happen when you click “see translation” beneath a post or comment—are now powered by neural networks, which are a form of artificial intelligence.

Back in May, the company’s artificial intelligence division, called Facebook AI Research, announced that they had developed a kind of neural network called a CNN (that stands for convolutional neural network, not the news organization where Wolf Blitzer works) that was a fast, accurate translator. Part of the virtue of that CNN is that instead of looking at words one at a time, it can consider groups of them.

Now, Facebook says that they have incorporated that CNN tech into their translation system, as well as another type of neural network, called an RNN (the R is for recurrent). Those RNNs, Facebook said in a blog item about the news, are better at understanding the context of the whole sentence than the previous system, and can reorder sentences as needed so that they make sense.

The upshot? Facebook says that the new AI-powered translation is 11 percent more accurate than the old-school approach, which is what they call a “phrase-based machine translation” technique that wasn’t powered by neural networks. That system translated words or small groups of words individually, and didn’t do a good job of considering the context or word order of the sentence.

As an example of the difference between the two translation systems, Facebook demonstrated how the old approach would have translated a sentence from Turkish into English, and then showed how the new AI-powered system would do it. The first Turkish-to-English sentence reads this way: “Their, Izmir’s why you said no we don’t expect them to understand.” Now check out the newer translation: “We don’t expect them to understand why Izmir said no.” Notice how the AI fixed the mistakes in word and phrase order?

While neural networks had been working together with the more traditional translation system before today, now all the translation gets its smarts from AI. This new system is capable of translating in 2,000 “directions.” For example, a translation from English to French is one direction, French to English is a second, and French to Italian is a third direction, and so forth. Astoundingly, the neural networks handle 4.5 billion translation per day, making them quite the linguists.

This smart window uses electricity to quickly change from clear to dark

In a world where even mundane devices like water bottles and toothbrushes have become smart and connected, it’s easy to scoff at yet another attempt at smartifying practical gadgets. But new technology for smart windows, described today in the journal Joule, might actually be an intelligent idea. That’s thanks to their energy-saving potential: Dynamic windows that darken when the sun is shining on them could help reduce cooling costs in the summer, just like putting down the blinds does. And with the help of an internet connection and an intelligent schedule, these could be automatic, so you wouldn’t have to remember when to turn them on and off.

Windows that tint on demand already exist—one prominent class of them is known as electrochromic windows. In fact, if you’ve flown on a Boeing 787 Dreamliner, you’ve seen electrochromic windows on the fuselage, which dim with a button’s press, either by the passenger or the flight crew. But a group of researchers at Stanford are aiming for a better, more dynamic one: Their prototype goes from transparent to opaque in less than three minutes. And it does it using an innovative approach.

Electrochromic technology in general isn’t perfect, says Ioannis Kymissis, an associate professor of electrical engineering at Columbia University. “They’re not amazing, but they’re not terrible,” he says. “For privacy applications, they’re not as high performing as people would like.” In other words, they might not get dark enough to be totally opaque, and the transition time can be slow. (One brand, SageGlass, says their windows take between 7 to 12 minutes to transition.)

But this new approach is clever, says Kymissis, who wasn’t involved in the new research. Their method involves using ions from two metals, like copper and silver, in an electrolyte gel on the window. The glass also needs to have indium tin oxide in it, a transparent conductor that’s ubiquitous in television and smartphone screens. By applying a negative electrical voltage, the window becomes dark because the ions form elemental, solid metal, which is opaque. A positive voltage causes the metal to dissolve back into ion form, allowing the light to come through.

It’s all fun and games until robots take over

Good morning fellow campers! It’s another beautiful Monday here at Camp PopSci as we once again transition from warm sunshine to the familiar glow of our computer screens. Last week wasn’t big on hardware announcements—that’s expected in these warm summer months—but there’s still plenty of interesting stuff to check out before heading off to the arts and crafts tent to make an ashtray or macaroni necklace. See you all in the mess hall for lunch!

This week’s Musk-Read

Last week’s camp champion was camper OpenAI, an artificial intelligence bot backed by Elon Musk. The robotic gamer was able to beat some of the best players in the world at the super-popular and wildly complex video game, Dota 2. The conditions for the robotic victory still have to be pretty specific—it can only play as and against a specific character—but the learning capabilities seem very impressive. Elon showed some team spirit via Twitter about the victory.


Summer romances come to an end—hopefully with a touching musical number—and now Netflix and Disney are officially looking to break up. Disney’s plan will tentatively remove all its content from Netflix and start its own streaming properties by 2019. There are reportedly still talks to keep Marvel content on Netflix, but the future is unclear at the moment.

For the records

The Internet Archive now plays host to more than 50,000 digitized recordings from 78 rpm records and even cylinders from the Archive of Contemporary Music. The music isn’t tidied up or remastered, so the recordings have audio evidence of the physical imperfections on the original media. It may not be to your taste, but at least it’ll give you a break from hearing “Despacito” on loop.

The sound of silence

Every single time I plug my phone into my car’s USB port, it plays “Abracadabra Holmes” by Daggermouth at a volume that’s somehow always too loud. It’s not a song I particularly love, but it’s impossible to stop it from autoplaying, or at least it was until now. A former Buzzfeed employee released a song called “A a a a a Very Good Song,” which is actually roughly 10 minutes of silence. Buy it, and it will become the first song in your iTunes library when listed alphabetically, so it will auto play, granting you enough time to pick a song you actually want or turn off the sound completely.