TechSummit Rewind 161

Netflix ditches five-star ratings, New figures on Uber’s anonymous division, and Qualcomm rebrands its processors

This is the TechSummit Rewind, a daily recap of the top technology headlines.

Netflix ditching five-star ratings in favor of thumbs up

Earns Netflix

Netflix will soon make its first change to ratings in years, switching from a traditional five-star rating to a binary thumbs up/thumbs down system, per a press briefing from company vice president of product Todd Yellin.

“Five stars feels very yesterday now. We’re spending many billions of dollars on the titles we’re producing and licensing, and with these big catalogs, that just adds a challenge.

“Bubbling up the stuff people actually want to watch is super important.”

-Todd Yellin, Netflix VP of product

Per Yellin, the change will happen next month globally.

“What’s more powerful: you telling me you would give five stars to the documentary about unrest in the Ukraine; that you’d give three stars to the latest Adam Sandler movie; or that you’d watch the Adam Sandler movie ten times more frequently? What you do versus what you say you like are different things.”

-Todd Yellin, Netflix VP of product

In addition to the ratings change, Netflix will also start percent matching, meaning that it will use algorithms to show a percentage below a title based on how likely it is a viewer will enjoy it. This is personalized, per Yellin, like dating sites that match you with potential partners based on interests or earlier activity.

Netflix is also “matching” members based on a global database of activity, not segmenting it by local markets. The company found that its members are willing to watch Netflix content that has been produced in other countries or has subtitled.

“We’re finding these clusters of people and then we’re figuring out who is like you, who enjoys these kinds of things, and then we’re mixing and matching those.”

-Todd Yellin

Docs: Uber’s anonymous cars drove over 20K miles, had to be taken over at every mile


Per documents circulating through Uber’s self-driving group obtained by Recode, the company’s 43 active cars in Pennsylvania, Arizona, and California drove 20,354 miles autonomously last week for only the second time since late December.

Uber passengers took around 930 self-driving rides in Pittsburgh last week and around 150 rides in Phoenix. These vehicles had a driver at the wheel to take over if needed.

However, those human drivers are taking over more often than they did in January.

Uber uses several methods to determine how its systems have progressed. Those include:

  • The average number of miles a car drives itself before a driver must take over for any reason
  • The average number of miles between “critical” interventions (when a driver must avoid causing harm, like hitting pedestrians or causing material property damage)
  • The average number of autonomous miles between “bad experiences” (jerky motions or hard braking, which are more likely to cause discomfort than damage)

During the week that ended Mar. 8, the 43 active cars on the road drove only an average of close to 0.8 miles before the safety driver had to take over for some reason.

The mile per intervention metric includes all the times driver have had to take back control from the system over the course of a week.

Reasoning for these interventions can vary, but they can include navigating unclear lane markings, the system overshooting a turn or driving in inclement weather. This excludes “accidental disengagements, end-of-route disengagements, and early takeovers.”

That’s down slightly from earlier this year. At the end of January, a driver had to take over roughly once every 0.9 miles and was at the one-mile mark during the first week of February.

Then there’s the company’s “critical” interventions. Last week, the company’s cars drove an average of approximately 200 miles between those types of incidents that required a driver to take over.

While that’s an improvement from last week, which was about 114 miles between critical interventions, that progress hasn’t been steady.

At the end of January, drivers only needed to take over after an average of 125 miles driven, but that dropped to about once per 50 miles during the first week on February. Those numbers then increased over the following two weeks but dropped again in the first week of March.

Part of that can be blamed on the cars being introduced to new routes (parts of Arizona) or having to navigate around objects or road markings they don’t recognize.

The cars also had more “bad experiences” during the week ending on Mar. 8 than in January. The miles driven between things like auto-detected hard decelerations or abrupt car jerks and movement has been cut in half from over four miles in January to less than two miles last week.

Per Uber’s self-driving team, the rider experience dropped significantly along Arizona’s Scottsdale Road. Cars were only able to drive 0.67 miles between interventions and two miles between bad events.

Google Home plays Beauty and the Beast audio ads


Some Google Home owners have reported hearing an advertisement for the opening of Beauty and the Beast alongside a summary of the day ahead.

Some Android users also reporting hearing the ad through Google Assistant.

The ad was delivered with the usual Google Assistant voice, so it blended in seamlessly with the usual daily briefing of news, weather, calendar appointments, etc.

The company has since stopped serving that content to users.

“This wasn’t intended to be an ad. What’s circulating online was a part of our My Day feature, where after providing helpful information about your day, we sometimes call out timely content. We’re continuing to experiment with new ways to surface unique content for users and we could have done better in this case.”

-Google, in a statement

Qualcomm Snapdragon brand changing to reflect ‘platform’ capabilities


Most of the mobile devices we know and love run on Qualcomm Snapdragon processors that often get mistaken as just CPUs.

For that reason, Qualcomm is subtly changing its branding and messaging to now call the processors the “Qualcomm Snapdragon platform.” This helps Qualcomm explain that it’s more than just a processor inside – instead, it’s a system-on-a-chip with a cellular modem, GPU, and more.

“We can now articulate the value that we provide to a device manufacturer – from developing algorithms for great pictures and videos, to making sure that the battery is long lasting. More importantly, the word “platform” will be used to explain the combined key user experiences – camera, connectivity, battery life, security, immersion – that these essential technologies are designed to deliver.”

-Qualcomm, in a statement

This signals the chipmaker’s movement into selling its wares to more than just phone makers – think automotive, IoT, and laptop – with a broader branding paintbrush.

As part of this change, the Snapdragon name is being removed entirely from its low-end chips. The current Snapdragon 200 range will now be known as “Qualcomm Mobile.”

iMessage App Store growth slows

iMessage App Store

The excitement is fading for the iMessage App Store along with its growth. During its first few months of existence, the store saw growth of over 100 percent month-over-month. Between January and February, that’s dropped down to just nine percent.

Per a new report from app intelligence firm Sen.sor Tower, there are nearly 5,000 iMessage-enabled apps (the same number of iOS apps released in year one of its App Store).

Games continue to be the most popular iMessage app category, ahead of entertainment, utilities, social networking, and photo & video apps. However, even within these categories, many of the apps are stickers – for example, those that use existing IP from a popular gaming franchise.

Gmail can now stream video attachments on desktop


Desktop Gmail users can now stream video attachments on the page, rather than being forced to download it first.

Attachments can still be downloaded, but clicking on a file will now pull up a YouTube-like video player that’ll let you play the clip back, adjust quality and volume levels, and even stream it to a Chromecast.

Per Google, the feature will roll out to everyone over the next 15 days.

Swatch launching smartwatch OS

A Swatch Scuba Playero wrist watch is displayed in a shop in Zurich

Swatch is developing an alternative to watchOS and Android Wear.

The company’s Tissot brand will launch a smartwatch towards the end of next year with the OS built in, per Swatch CEO Nick Hayek. Hayek claims that the system will need less power and will better protect data.

“There’s a possibility for wearables to develop as a consumer product, but you have to miniaturize and have an independent operating system.”

-Nick Hayek, Swatch CEO

Swatch is willing to give third parties access to the operating system co-developed with the Swiss Center for Electronics and Microtechnology, per Hayek. The company has received about 100 requests for more information, with half coming from smaller Silicon Valley companies.

Amazon puts Alexa inside main iPhone app


Available to all iPhone users next week, Amazon iOS app users will be able to talk to the company’s Alexa assistant.

Naturally, the assistant can shop and track packages, but she can also do other quirky things like tell jokes, give weather updates, and predict items like Best Picture at the Academy Awards or the winner of the NCAA Basketball Tournament. It also plays music, controls Internet of Things devices, and grants Amazon app users access to over 10,000 skills.

At launch, Alexa won’t support the Door Lock API that lets users lock (and eventually unlock) doors with smart locks.

Settings changes still require use of the separate iOS Alexa app.

Nvidia partners with PACCAR on self-driving truck tech


Nvidia has launched a new autonomous vehicle partnership with PACCAR, one of the largest makers of transport trucks.

The arrangement has already given us one proof-of-concept vehicle, a Level 4 autonomous truck that uses Nvidia’s Drive PX 2 platform with neural network training fed by data of humans driving tractor-trailers. The partnership’s announcement included a showcase of this initial vehicle managing a closed road course, with no one behind the wheel.


TechSummit Rewind #066: Google I/O recap

Editor’s Note: This is a special edition of the TechSummit Rewind, a daily recap of the top technology headlines.

Google I/O is the Internet giant’s biggest showcase of everything going on inside Mountain View, and boy did they take advantage of that spotlight. Here’s what was announced:

Google Assistant

Google Assistant is a virtual assistant that does everything you know and love from Google Now and other voice assistants, but with a conversational touch.

That touch is being offered in two ways: Google Home, an Amazon Echo-esque speaker, and chat app Allo. The assistant is customized based on the platform it’s on.

Google Home is a small speaker that lets you converse with it in natural language, and gives Google a token into your everyday life. It’ll be available this fall.

If this seems a lot like Amazon’s Echo, that’s because it is. In a video launched Wednesday, Google laid out its vision for the product: where people can communicate with Home to do everything from play music, check the weather, change dinner reservations, and send texts.

In the case of Allo, the free app will be available later this summer on iOS and Android.

The app lets you sign up with your phone number, with your Google account connected to it if you choose.

When you send photos, they show up full-bleed on screen with the ability to doodle on them. You can also drag your finger up/down on the send button to enlarge or shrink text in a feature Google dubs WhisperShout.

However, Google Assistant is the star of the show. You can set up a conversation with Google (@google) and ask it questions. The assistant will respond with the information you’d expect from the search engine, but with a conversational touch. It’ll suggest further searches, and give you ways to take action through Google.

Allo’s chatbot is smarter than others, thanks to the power of Google’s Knowledge Graph, which allows you to ask more complex questions that can’t be answered by simply crawling the web. If you get bored, @google can also play a game like “guess the movie based on a string of emoji.”

Inside conversations with friends, Allo will put suggested replies at the bottom of new messages powered by a machine learning engine. For example, a graduation photo will spur lines like “Congratulations!” and “You look great!” These responses are grouped into meaning clusters, giving you a range of possible reactions. These chips are based on machine learning applied to the way you type.

When you hit one of these suggestions or type @google, everyone in your conversation can see and respond to the answers.

In terms of privacy, all conversations are encrypted “on the wire,” meaning that nobody online can read them as you send your messages. They are read by Google’s servers, but aren’t kept around and have no identity attached to them. If that’s not strong enough for you, there’s also an Incognito Mode – similar to the Chrome mode of the same name. When you enable it, your conversation is subject to end-to-end encryption, with Google not even having the ability to read it. Notifications for Incognito chats also don’t reveal their contents on the lock screen, ensuring a high level of privacy.

Google Duo speeds up mobile video calls

Alongside Allo, Google is launching the Duo video calling app for iOS and Android.

The app greets you with a selfie-cam video preview, and giant circled photos of contacts that can be used to start a call – scrolling up will reveal more contacts. When you call, the person on the other end answers, and a video chat begins. You can mute your microphone or flip the camera, but you can’t video conference.

The kicker is a live video preview of the person you’re calling when receiving one on your lock screen. When you answer, the video call has already started.

Under the hood, Google has spent time optimizing for speed and latency to work even on flakier connections, with audio/video quality adjusting to fix the connection. Duo can also seamlessly handle switching from WiFi to cellular data without dropping the call.

Android N

Google has launched a website to let you suggest a name for Android N, with Google using those suggestions to pick a final name.

The OS’ multitasking menu has been streamlined to be limited to the last seven apps used, because that’s what most people jump back too, along with adding a “Clear All” button to dismiss all active apps.

However, the biggest addition comes in the form of a VR mode that’s part of Google’s wider Daydream initiative. Phones that adhere to a “Daydream Ready” spec will come from all the major Android manufacturers (Samsung, LG, Huawei, HTC) this fall. The company has also secured content partnerships with HBO, Netflix, Hulu, CNN, the NBA, the Wall Street Journal, the New York Times, and IMAX.

A beta version of Android N is available for download today for Nexus devices and the Google Pixel C tablet.

Lastly, there’s Instant Apps, which allows Android users to use apps from a link without having to download it from the Play Store. Google pitched it was a way to use single purpose apps, like paying for museum parking, without having to deal with the mobile web or having another app clutter your homescreen. After clicking a link, Google Play downloads the parts needed for the desired experience. Users will then have the option to download the full app.

Instant Apps will be compatible with devices running Android 4.1 Jelly Bean or later when it launches later this year.


In addition to using a headset, Daydream also features a new control system for mobile VR.

The headset is a step up from Cardboard, with a front slot made to hold a smartphone that latches at the top.

That control system comes in the form of a remote with volume buttons and a round pad on top, similar to the Oculus Touch controller that ships with the Rift headset. These design also features basic motion sensing capabilities.

Android Wear

Android Wear 2.0 is the biggest update to the wearable OS since it launched in September 2014. The headline feature is the ability for apps to run standalone without the need for a nearby phone. This’ll help Android Wear focus on fitness, messaging, and customization. For instance, watch faces can show data from any app now, similar to the Apple Watch’s complications.

Conversing via text with the watch is now easier, with better automatic exercise recognition, sync with third-party apps, and exchange data via Google Fit.

Chrome OS

The Google Play Store is on its way to Chrome OS users.

Starting in early June, Android apps will be available to touchscreen-toting Chromebooks like the Asus Chromebook Flip, 2015 Chromebook Pixel, and Acer Chromebook R11.