Google’s search engine for scientists upgraded for better data scouring

Google’s search engine for datasets, the cunningly named Dataset Search, is now out of beta, with new tools to better filter searches and access to almost 25 million datasets.

Dataset Search launched in September 2018, with Google hoping to slowly unify the fragmented world of online, open-access data. Although many institutions like universities, governments, and labs publish data online, it’s often difficult to find using traditional search. But by adding open-source metadata tags to their webpages, these groups can have their data indexed by Dataset Search, which now covers a huge range of information — everything from skiing injuries to volcano eruptions to penguin populations.

Google would not share any specific usage figures for the search engine, but it said “hundreds of thousands of users” have tried Dataset Search since its launch, and the reaction from the scientific community was overall positive.

Natasha Noy, a research scientist at Google AI who helped create the tool, tells The Verge that “most [data] repositories have been very responsive” and that the engine’s launch meant older scientific institutions are now taking “publishing metadata more seriously.”

“For example, [the prestigious scientific journal] Nature is changing its policies to require data sharing with proper metadata,” Noy says, highlighting a change that will make the data underpinning top-flight scientific research more accessible in future.

“Finally! My thesis ‘Hitting The Slopes A Little Too Hard: Shattered Femurs and Broken Dreams In the 2012 World Ski Cup,’ will have the rigorous, data-based grounding it deserves.”
Image: The Verge

New features added to Dataset Search include the ability to filter data by type (tables, images, text, etc), whether it’s free to use, and the geographic areas it covers. The engine is also now available to use on mobile and has expanded dataset descriptions.

Google says the corpus covered by the search engine — almost 25 million datasets — is only a “fraction of datasets on the web,” but a “significant” one all the same. The largest topics indexed are geosciences, biology, and agriculture, and the most common queries include “education,” “weather,” “cancer,” “crime,” “soccer,” and “dogs.” The US is also the leader in open government datasets, publishing more than 2 million online.

Noy would not comment on future plans for Dataset Search, but she says the team was thinking about a number of functions they hope would be useful, including “understanding how datasets are cited and reused” and “helping users explore datasets in Dataset Search when they don’t necessarily know what they are looking for.”

“And, of course, continuing to expand the corpus,” says Noy. There’s always more data out there.

Waymo is bringing its self-driving minivans and trucks to Texas and New Mexico

Waymo is hitting the open road and bringing some of its self-driving minivans and long-haul trucks to two new states. The Alphabet subsidiary said in a tweet on Thursday that it would deploy a portion of its fleet in Texas and New Mexico as it seeks to learn about new road conditions and environments. It also feeds into Waymo’s ultimate quest to gain a toehold in more markets for the eventual launch of a commercial business.

Waymo didn’t specify in which cities it will start testing its vehicles, nor did it explain what type of testing it will conduct. For example, last year, Waymo brought three of its self-driving Chrysler Pacifica minivans to Los Angeles but only to map the downtown area.

It’s especially notable that Waymo’s autonomous tractor-trailers are being deployed in two new states. The company has previously only tested its Class 8 trucks in California, Arizona, and Georgia, so the territorial expansion is sure to be good news for Waymo’s stated plans to eventually launch a commercial freight hauling business.

“These are interesting and promising commercial routes, and we’ll be using our vehicles to explore how the Waymo Driver might be able to create new transportation solutions,” the company says. (“Waymo Driver” is the company’s shorthand for its autonomous vehicle hardware and software system.)

The company isn’t a complete stranger to the great state of Texas. Waymo’s groundbreaking demonstration of its prototype Firefly vehicle with no steering wheel or pedals took place in Austin in October 2015. The company kept an office in the city until November 2019 when it abruptly shut it down. A reported 100 employees and contract workers lost their jobs as part of the decision, according to CNBC.

Perhaps more intriguingly, though, was a sneak peek at the company’s next-generation hardware suite in a tweet from Waymo CEO John Krafcik. The photo shows Krafcik along with Axios reporter Joann Muller standing in front of a Jaguar I-Pace electric vehicle with camouflaged sensors on the roof.

In 2018, Waymo inked a deal with Jaguar Land Rover to purchase 20,000 I-Paces to be self-driving taxis. At the time, they said the new vehicles would officially become part of Waymo’s commercial ride-hailing service starting in 2020. The company is already testing the new vehicle on public roads around its Mountain View headquarters.

With the electric SUVs primed to join the fleet later this year, there’s no better time to start showing off the latest member of the autonomous family. Waymo is expected to have more say about its fifth-generation vehicle in the months to come.

Google’s ads just look like search results now

Last week, Google began rolling out a new look for its search results on desktop, which blurs the line between organic search results and the ads that sit above them. In what appears to be something of a purposeful dark pattern, the only thing differentiating ads and search results is a small black-and-white “Ad” icon next to the former. It’s been formatted to resemble the new favicons that now appear next to the search results you care about. Early data collected by Digiday suggests that the changes may already be causing people to click on more ads.

The Guardian’s Alex Hern is one of many commenters to point out the problem, noting that there’s now next to no visual distinction between ads and search results. “There is still, technically, *labelling*, but it’s hard to escape the conclusion that it is supposed to be difficult to spot at a glance where the adverts end,” he tweeted.

It’s especially striking considering how distinct Google designed its ads in the past. Up until 2013, the search engine gave its ads an entirely different background color to distinguish them from its organic search results. But even after that, it continued to use unique colors that effectively let users quickly see where its ads ended and organic results began.

In the past, Google used to give its ads a different background color to make them distinct from the rest of its results. This screenshot is from 2013.
Image: Search Engine Land

In a blog post announcing the new design when it came to mobile last year, Google partially explained the change by saying that adding favicons to organic search results means that “a website’s branding can be front and center,” which means “you can more easily scan the page of results.” But it spent far less time talking about the changes to its ad designs, which now feel much more significant, especially when viewing results on a laptop or monitor.

In the past, Google’s Sundeep Jain justified simplifying the company’s ad designs by saying that a simpler design “makes it easier for users to digest information,” according to Search Engine Land. He added that the company was trying to reduce the number of different colors used on a page in order to bring a little more “harmony” to the layout.

It’s hard not to get the feeling that this “harmony” is less about offering a better user experience, and more about helping Google’s ad revenue. As Digiday reports, there’s data to suggest that’s actually the case. According to one digital marketing agency, click-through rates have already increased for some search ads on desktop, and mobile click-through rates for some of its clients increased last year from 17 to 18 percent after similar changes to Google’s mobile search layout.

Google is fundamentally an ad business. In the third quarter of 2019, Google’s parent company Alphabet made nearly $34 billion from Google advertising, out of a total revenue of $40 billion for Alphabet as a whole. At that sort of scale, small changes in ad click-through rates could end up having a huge effect on Alphabet’s bottom line, even if it means tricking users for cheap clicks.

This is Microsoft’s vision for dual-screen apps on Windows 10X and Android

Microsoft is starting to share more details on exactly how it imagines apps will run on dual-screen devices like the Surface Duo and Surface Neo. The software giant unveiled both devices back in October, with the smaller Surface Duo running Android and the larger Surface Neo powered by Windows 10X. Now, Microsoft is getting developers ready to test their apps to see how both devices will span them across both displays.

By default, an app will occupy a single screen according to Microsoft. Surface Duo or Surface Neo users can then span the app across both displays when they’re in double-portrait or double-landscape layout. Microsoft envisions that app developers will experiment with different ways to utilize both screens. Some of these include simply using both screens as an extended canvas, having two pages of a document shown at once, using the second display as a companion or dual view of something, or having a master part of the app on one display and details on the second.

Microsoft’s dual-screen app vision.
Image: Microsoft

These are “initial app pattern ideas,” according to Microsoft, and the company could well extend them based on developer feedback in the coming months. Microsoft is also asking developers to consider a variety of inputs and to embrace the way orientations these new dual-screen devices will be used in.

On the input side of things, Microsoft wants developers to optimize apps that are built for things like note taking to work best in landscape mode. “Our studies show that users are more comfortable typing or writing on a flat surface,” says a lengthy Microsoft dual-screen document. These variety of inputs include stylus, pen, and touch, just like many Windows 10 devices today.

While Windows 10 already natively supports touch, Microsoft wants dual-screen app developers to embrace drag and drop. Because apps open fullscreen on one display by default, you can also window them to have two running side-by-side on each display. “A dual-screen device that runs apps side-by-side especially lends itself to drag-and-drop interactions for a great app experience,” explains Microsoft. If developers enable multiple instances for their apps, you’ll also be able to open multiple instances of the same app and run them side-by-side.

Microsoft appears to be readying developers largely for dual-screen devices, and not single foldable displays. These multiple displays do introduce an obvious seam in the middle, and Microsoft is providing some guidance on embracing that gap or avoiding it. Apps like Google Maps can embrace the seam easily because map data can be moved freely, but some apps will have dialog boxes that were designed to be centered, or images that are displayed in the center.

Microsoft is recommending that app-launched dialogs should move to one screen instead of trying to stretching across both. That’s the same for context menus or drop down menus, that should ideally treat the gap between the displays as a boundary and not try to display menus there.

Developers will also need to consider snapping to the seam, to avoid content in a grid or tabular layout not being displayed correctly. Responsive layouts will help here, by rearrange content when the screen orientation or size changes. Developers can also choose how to handle images spanning across the two displays. They can either mask, where an image is rendered behind the seam, or split which cuts the image apart. The mask technique leaves it up to a users’ brain to “naturally connect the ‘unseen’ parts,” according to Microsoft, and it’s a recommended technique for videos and photos.

Microsoft is also releasing an Android emulator for the Surface Duo today to allow devs to test mobile apps. A Windows 10X emulator for the Surface Neo will arrive next month at around the same time that Microsoft plans to detail more of its dual-screen plans during a developer webcast.

Microsoft’s Android emulator will naturally support Android apps, and the Windows 10X version will include support for native Windows APIs to let developers detect hinge positions and optimize their win32 or Universal Windows Platform (UWP) apps for these new devices. Microsoft is also proposing new web standards for dual-screen layouts, and is “actively incubating new capabilities that enable web content to provide a great experience on dual-screen devices.”

Windows 10X Emulator.
Image: Microsoft

Other OEMs like Dell, HP, Lenovo, and Asus are also working on Windows 10X dual-screen and foldable devices. We’ve already seen one from Lenovo in the form of the ThinkPad X1 Fold, but we’re hoping to see more in the coming months. Microsoft is also planning to reveal more details about its dual-screen plans at the company’s Build developer conference in May.

Update, January 23rd 4:20AM ET: Article updated with more information from Microsoft’s dual-screen documentation.

Google publishes largest ever high-resolution map of brain connectivity

Scientists from Google and the Janelia Research Campus in Virginia have published the largest high-resolution map of brain connectivity in any animal, sharing a 3D model that traces 20 million synapses connecting some 25,000 neurons in the brain of a fruit fly.

The model is a milestone in the field of connectomics, which uses detailed imaging techniques to map the physical pathways of the brain. This map, known as a “connectome,” covers roughly one-third of the fruit fly’s brain. To date, only a single organism, the roundworm C. elegans, has had its brain completely mapped in this way.

Connectomics has a mixed reputation in the science world. Advocates argue that it helps link physical parts of the brain to specific behaviors, which is a key goal in neuroscience. But critics note it has yet to produce any major breakthroughs, and they say that the painstaking work of mapping neurons is a drain on resources that might be better put to use elsewhere.

“The reconstruction is no doubt a technical marvel,” Mark Humphries, a neuroscientist at the University of Nottingham, told The Verge. But, he said, it’s also primarily a resource for other scientists to now use. “It will not in itself answer pressing scientific questions; but it might throw up some interesting mysteries.”

The 3D map produced by Google and the FlyEM team at Janelia is certainly a technical achievement, the product of both automated methods and laborious human labor.

The first step in creating the map was to slice sections of fruit fly brain into pieces just 20 microns thick, roughly a third the width of a human hair. Fruit flies are a common subject in connectomics as they have relatively simple brains about the size of a poppy seed but display complex behaviors like courtship dances.

These slices of brain are then imaged by bombarding them with streams of electrons from a scanning electron microscope. The resulting data comprises some 50 trillion 3D pixels, or voxels, which are processed using an algorithm that traces the pathways of each cell.

Despite Google’s algorithmic prowess, it still took substantial human labor to check the software’s work. The company says it took two years and hundreds of thousands of hours for scientists at Janelia to “proofread” the 3D map, verifying the route of each of the 20 million chemical synapses using virtual reality headsets and custom 3D editing software.

Even then, the resulting map only covers a portion of the fruit fly’s brain, known as the hemibrain. In total, a fruit fly’s brain contains 100,000 neurons, while a human brain has roughly 86 billion. That suggests how far we are from creating a full connectome of our own neural pathways.

Joshua Vogelstein, a biomedical engineer and co-founder of the Open Connectome Project, told The Verge that the work would be a boon to scientists. Vogelstein said that in the decade to come, the data provided by such projects would finally start to yield results.

“I believe people were impatient about what [connectomes] would provide,” said Vogelstein. “The amount of time between a good technology being seeded, and doing actual science using that technology is often approximately 15 years. Now it’s 15 years later and we can start doing science.”

Google and the FlyEM team have made the data they collected available for anyone to view and download. The group has also published a pre-print paper describing their methodology, and say they’ll be publishing more papers on their work in the weeks to come.

Google designed an envelope you can use to hide your phone from yourself

Today, Google launched three new experimental apps to help you use your phone less as part of its Digital Wellbeing initiative, including one that actually has you seal up your phone in a phone-sized paper envelope (via Android Police). It sounds similar to the pouches some artists require fans to put their phones into at concerts, except it’s something you make at home — and Google’s envelope should at least let you make a call, if you need to.

If you want to try Google’s envelope technique out for yourself, it only works with a Pixel 3A right now — and we can’t think of a reason that’s the case. If you do have a Pixel 3A, though, download the required Play Store app for the envelope, called Envelope, then print out the PDF for the envelope, cut out the template, and follow the instructions to construct it.

Then, when you’re ready for a break from your phone, open up the Envelope app, slide your Pixel 3A into the envelope, and seal the envelope shut — the PDF recommends using glue. Once your phone is sealed in the envelope, you’ll only be able to dial phone numbers on the phone, use speed dial, or have the phone tell you the time by flashing the numbers on the number pad.

Check out this video to see how it all works:

In the video, Google also shows off a version of the envelope that only lets you take photos and videos, but I haven’t found a PDF of that. Here’s a better look at it:

Image: Google

Image: Google

I actually really like the idea of Envelope. I’ve been to a couple of those concerts that required using the phone pouches, and I felt that being completely blocked from using my phone helped me and the rest of the crowd focus better on the show. I can see how Envelope could be a similarly useful way to give myself a break from my phone at home. But since your phone is glued inside the envelope, you’ll have to tear the seal and part of the envelope to get it out, which means that you’d probably have make a new envelope every single time you want to use the app.

Google has open sourced the Envelope app on GitHub, so if you want to make an envelope that lets you do something different than call people or use the camera, you can tinker what Google has already shared to your liking.

Google also launched an Android app called Activity Bubbles, which puts a bubble on your wallpaper each time you open the phone and increases the size of the bubble the longer you’re on your phone. If your home screen is covered in bubbles, that might be a sign that you’ve been on your phone for too long.

And if you want to visualize your time on your phone in hard numbers, consider Google’s new Screen Stopwatch, a live wallpaper for Android phones that constantly counts up whenever you have your phone unlocked. These screenshots from Google show how the numbers add up over time:

If you’re looking for other ways to evaluate your phone use, late last year, Google released a handful of other Digital Wellbeing experiments on the Play Store, such as one that counts how many times you unlock your phone in a day.

Netflix uses Google Trends to claim The Witcher was more popular than The Mandalorian

Netflix released its quarterly earnings report Tuesday afternoon, including a trove of new data on its latest hit, The Witcher. Among the data points Netflix cites in promoting its business over the last three months includes a Google Trends chart comparing interest in its new fantasy adaptation to Disney+’s The Mandalorian, the other huge streaming hit of the holiday season.

According to Netflix, The Witcher blew the competition out of the water, including trouncing not only the Star Wars spinoff, but also Apple TV Plus’ The Morning Show and Amazon’s Jack Ryan. Yet the data is a bit misleading. Netflix’s included chart relies on global Google search data. That’s an issue when comparing Netflix hits to those of new streaming services from Apple and Disney, both of which only launched in the last few months.

Screenshot by Nick Statt / The Verge

As you can see above, a popular Netflix show made available all across the world naturally trounces programs made available only in a few territories. That makes sense. Netflix is available in 190 countries, and The Witcher — based both on a popular Polish book series and its video game counterpart from a Polish game studio — was made available in multiple languages, helping its global appeal.

Disney+, on the other hand, has only been available in five countries since its November launch: the US, Canada, the Netherlands, Australia, and New Zealand. So although The Mandalorian featured more language and subtitle support than The Witcher, it simply can’t be watched in anywhere near as many countries.

If we adjust the Google Trends chart from “worldwide” to “United States,” we get a more telling look at the two hit shows’ popularity in comparison:

Screenshot by Nick Statt / The Verge

What we see here is emblematic of the difference in release approach between Disney+ and Netflix, and the results are not as flattering for Netflix as they may seem. The Witcher did indeed enjoy a huge spike in searches with its release in mid-December. That’s to be expected, considering Netflix drops entire seasons all at once nearly everywhere its service is available. And yes, that initial spike was greater than any one moment for The Mandalorian.

But Disney has opted for a traditional, weekly release approach in line with the standard TV practice we see HBO and others keep using. And it looks like it paid off. The Mandalorian experienced sizable, sustained boosts from its entire eight-episode run over the course of two months. The finale and the premiere both had similar spikes in line with The Witcher’s popularity just a week or so after its release, indicating The Mandalorian likely maintained an impressive level of popularity throughout its entire run. Right now, The Witcher as a whole is less popular than at any time a new episode of The Mandalorian released throughout its first season.

As my colleague Chaim Gartenberg noted last month, this clashing of release strategies may help Netflix in the short term, but hurt the platform in the long run: “By dropping every episode at once, Netflix is sacrificing weekly discussions around The Witcher for a short burst of popularity, after which it trickles off into the void as people’s attentions are quickly grabbed by the next big thing. That extra time between episodes would let viewership build over time, as more people hear about the show or proselytize it to their friends.”

The Mandalorian, on the other hand, had the benefit of a enormous, organic meme in the form of Baby Yoda on top of a weekly and passionate deluge of recaps and analysis, fan theorizing, and genuine discussion on the series and its many Easter eggs and other Star Wars lore references. Perhaps this isn’t as big of a deal for Netflix right now, especially if its big hits keep clocking impressive viewing metrics like The Witcher did.

But if the goal for all of these services is to create the next big, buzzy hit show in the post-Game of Thrones era, the all-at-once release method may not be the best approach. And the Google Trends data proves it.

Google’s Pixelbook Go Chromebook is now available in that pink color

Google’s Pixelbook Go Chromebook is available in a color other than black now that the “not pink” version of the laptop is shipping from the US Google Store. Although the base $649 configuration with an Intel Core M3 chip is still only available in black, you can get the pink model if you’re willing to step up to the $849 Core i5 version we tested in our review. Chrome Unboxed was first to spot the availability of the new color scheme.

We can’t imagine there are many people who’ve been holding off on buying a Pixelbook Go purely because of the limited color options. Nevertheless, it was a decent-enough laptop when we tried it out for ourselves. Battery life is good, the keyboard is excellent, and overall it’s a solidly built laptop. Its slightly high price means that we wouldn’t go as far as to recommend it as the best Chromebook around. For comparison, Acer’s Chromebook C871 was announced earlier today with a starting price of $329.

If you want to go all out, then there’s a version of the Pixelbook Go available with a 4K screen that was released in December for $1,399.

Vox Media has affiliate partnerships. These do not influence editorial content, though Vox Media may earn commissions for products purchased via affiliate links. For more information, see our ethics policy.

Huawei signs deal with TomTom for a Google Maps alternative

Chinese phone maker Huawei will work with Dutch digital mapping company TomTom to put maps on its phones, Reuters reports. A TomTom spokesperson told Reuters that the deal closed “some time ago,” but it became public late last week.

While TomTom maintains self-branded apps on iOS and Android, Reuters describes Huawei building its own apps with TomTom’s maps, traffic information, and navigation tools. TomTom has previously provided data for Apple Maps — it was part of a shambolic patchwork of data providers at launch, but Apple reportedly kept using its services after overhauling the app as well.

Previous reports have said Huawei is building a full-fledged mapping system known as “Map Kit.” That software would be meant for app developers and could use data from Russian tech giant Yandex and Huawei’s own “telecom base stations.” The TomTom deal could signal that Huawei is at least temporarily backing off Map Kit, or that it’s simply still working on the tech and needs a short-term solution.

Like many Android phone makers, Huawei has so far relied on Google Maps. But the Trump administration placed sanctions on the company last year, threatening its ties to American tech companies like Google. Huawei is still using Android for now, but the future of that deal is “unclear,” and Huawei is building its own operating system called HarmonyOS. The TomTom deal further reduces its dependance on Google — although mapping apps can be notoriously tough to get right, so Huawei isn’t in the clear yet.

It’s not just you: Google added annoying icons to search on desktop

Google added tiny favicon icons to its search results this week for some reason, creating more clutter in what used to be a clean interface, and seemingly without actually improving the results or the user experience. The company says it’s part of a plan to make clearer where information is coming from, but how?

To give you an idea of how minimal the change is, here’s what it looked like when Google made the same tweak last year to the browsing experience on phones:

In my Chrome desktop browser, it feels like an aggravating, unnecessary change that doesn’t actually help the user determine how good, bad, or reputable an actual search result might be. Yes, ads are still clearly marked with the word “ad,” which is a good thing. But do I need to see Best Buy’s logo or AT&T’s blue circle when I search for “Samsung Fold” to know they’re trying to sell me something?

Search results for “Galaxy Fold” are not clearer for having the favicons

The company tweeted that the change to desktop results were rolling out this week, “helping searchers better understand where information is coming from, more easily scan results & decide what to explore.” But though the logos have been visible in search results on Google’s mobile browser since last year, Google’s statement doesn’t address how successful or irrelevant the favicons might have been for mobile users.

When Google first launched, its sparse, almost blank search page and minimalist results were an extremely welcome change, compared to the detritus on other search home pages at the time (which persists on sites like Yahoo). Adding favicons makes Google’s search results look a little cartoonish, and if we think Facebook users who can’t determine a reputable news source from their racist uncle’s favorite blog are going to be assisted by tiny pictures on Google, well, we’re likely to be disappointed.

Google does often make changes to search that actually do improve user experience or results, though. In the past few months, Google changed its search algorithm so it doesn’t see a search query as a “bag of words,” improved its results to prioritize reputable news sources, and even added augmented reality results to searches.

If you’re intrigued by the new logos in your search results, Google provided instructions on how to change or add a favicon in search results for those who don’t know. Lifehacker also provided instructions on how to apply filters to undo the favicon nonsense and revert back to how the search results used to look. You can decide which how-to is the more useful.