Reddit launches a ‘News’ tab into beta testing

Reddit is rolling out its “news” tab into beta, the company announced this week. The expansion follows on Reddit’s initial test of a news-related feature that began this May, when an alpha version shipped to some users of Reddit’s iOS app. At the time, Reddit explained it wanted to give news its own dedicated home on its site in order to make it easier for those with a lot of subreddit subscriptions to find the news without having to hunt around.

To determine what’s newsworthy, Reddit says it first figured out which subreddits were engaging with news the most. It did this by looking at the most-clicked posts by domain in the subreddits. The company came up with a list of around 1,000 domains from media publishers focused on news. This list was used to help it surface those communities where news was regularly discussed.

To be clear, the domain list was only used to find the appropriate subreddits where news was often discussed – it doesn’t mean Reddit is limiting news stories it surfaces to those 1,000 domains.

In addition, Reddit has a few other requirements for the communities featured in the News tab. They must have active moderation, abide by Reddit’s content policies and its guidelines for healthy communities, and the community has to require the post title accurately reflects the article title.

The communities included in the News section of Reddit discuss a variety of topics, across business, science, sports, gaming, entertainment, tech news, and more – popular categories across Reddit as a whole. The majority of the posts are link posts, with the exception of some sports news where video is allowed.

The News tab itself then organizes the posts by category, so users can filter the news for themselves. And users can further configure the tab to their own liking.

While Reddit users are often known to actually break news by (sometimes unwittingly) being the first to spot things, the News tab is focused on showcasing the work from news publications.

It’s more of a scannable list of top stories with an active comments section. That’s something that you don’t find on a number of news sites these day, as many have removed commenting. Meanwhile, a lot of discussion around the news takes place on social media, like Facebook and Twitter – but it’s not necessarily centralized.

Given that the product is still in beta, Reddit is still listening to user feedback about the new feature.

As you’d expect, there’s a lot of it – from “this is a bad idea” to “can we see the list?” to “this site is such a joke” to “this is actually controlling the bubble” to “this sounds like a sales pitch to someone who has never used Reddit” to this feels like another ‘Facebook’ style change that nobody asked for” to completely unrelated complaints about other issues.

It could be that Reddit is hoping to attract some attention to its app in the wake of Facebook pulling Trending Topics – its news discovery feature – from its social network.

A larger audience of non-Reddit users may now be looking for another way to easily browse news on mobile – so Reddit thinks it may as well filter some of its existing content and pop it into a tab for easy access. But it’s not likely that people will turn to Reddit for news, especially when there are formidable alternatives like Google News and Apple News.

The feature is in beta for the time being. The company has not said when it’s publicly launching.


Source: Tech Crunch

Konsus looks to give companies a way to get specially designed documents in under a day

Fredrik Thomassen as a consultant used to have the resources to offload the annoying project tasks — like making PowerPoint presentations — but now that it’s gone, he and his team wanted to make that available for everyone.

Now the startup, called Konsus, wants to turn that around even faster. Konsus is a design marketplace where companies can quickly post design projects that they need for various parts of their jobs, like presentations, and designers can pick up those jobs and submit their work back — a task that could take up a lot of unnecessary time for an employee that might be better spent working on other parts of their job. Konsus said it is compressing that even further by now looking to provide a 12-hour turnaround for those companies. The company launched out of Y Combinator in 2016.

“[Employees] want to be valuable and spend time on core tasks,” Thomassen said. “The average knowledge worker, depending on various specifics, spends around 40 percent of that time on non-core tasks that should be outsourced. That’s the 40 percent we’re going after, and people quite readily understand it. Some companies have in-house design agencies and so on, and they are 3 or 4 times as expensive as we are, and they typically want to work on these larger or more grand projects and don’t want to work on the small projects that range from 10 hours to 15 hours. Most of the projects we do are these small, nominal projects that people would have had to do themselves.”

Konsus hires account managers and project managers handling the relationships with the customers to ensure that they’re getting the quality they need when they are posting projects like PowerPoint presentations onto the site. But Thomassen also said there are plenty of examples of those firms finding designers and contractors that they’ve decided to bring on full-time, and he’s fine with the startup being seen as a springboard for contractors that want to polish their skills for working with western clients — and even end up with a full-time job after that. A lot of the designers are coming in from eastern Europe, southeast Asia and other parts of the world that aren’t necessarily on the radar of these western firms.

Like many other modern services and marketplaces, Konsus hopes to come in at the bottom of a company and work its way up. One person or a team from a larger corporation will discover it, start using it and then eventually the startup might track that firm down and start talking about a custom team and dedicated emails. Then the outsourcers working for that firm goes through a background check, signs confidentiality agreements and goes through training on corporate branding material. Konsus’ revenue comes partly from subscriptions and people pre-paying to get a team, and the other half a pay-as-you-go model where firms get a rate and Konsus takes a commission.

“If you look at [big consulting firms], they have a similar solution as we have, and you can get support for all kinds of services — data entry, PowerPoint, various graphic design tasks — that make life much, much easier,” Thomassen said. “You go home from work and then you get it back in the morning, it becomes part of your workflow. That’s what we wanted to build for everyone else. Freelancers come to us from all corners of the world, they apply on our website, and we have our own recruiter work with them. We get around 5,000 to 10,000 people who apply, and we accept 10-20 depending on how many we need. The bar is extremely high.”

Of course, given that these are the kinds of tasks that firms might outsource without such a platform, Konsus has to potentially deal with larger consulting firms like Accenture, and there are plenty of startups looking to create an online labor marketplace that might not be targeting design just yet. But as those platforms start to put together a lot of potential customers, they’ll likely start asking for tools like Konsus — which means the company is going to have to figure out ways to outcompete early.

The company has raised $1.7 million from Sam Altman, the Slack Fund, Acequia Capital, Paul Buchheit, Geoff Ralston, John Collison and Liquid2 Ventures.


Source: Tech Crunch

Your second chance for Startup Battlefield at Disrupt SF

If you missed the deadline last week to apply for the renowned Startup Battlefield at Disrupt SF, have no fear. There is still one more chance at being part of the action!

Out of all of the early-stage companies exhibiting at Disrupt, three Startup Alley companies will have the opportunity to be selected as one of the “Wild Card” winners. You might be wondering, “what is Wild Card”? Wild Card is a Startup Alley exhibiting company that is selected by our TechCrunch editorial team to participate in the celebrated Startup Battlefield competition. This year we’re selecting three, and if you are exhibiting in Startup Alley, you could be one of the lucky winners.

Last year at Disrupt NY 2017, RecordGram got a table in Startup Alley, where they ended up being selected as one of the Wild Card companies and ultimately went on to win the Startup Battlefield competition — and took home $50,000. Guess what? This year, the grand prize of Startup Battlefield at Disrupt SF has been doubled to $100,000!

Plus, you’ll have all the benefits of being a Startup Alley exhibitor, including access to CrunchMatch, Disrupt’s matchmaking service between startups and investors. So far, the investors coming to Disrupt SF this year have investment funds in excess of $4 billion, and we’re expecting more to sign up in the coming weeks. Also, over the course of the three-day conference, there will be curated tracks of content across four unique stages in 12 different verticals, plus tons of educational workshops and a plethora of networking opportunities.

So, if your company is pre-series A, Startup Alley at Disrupt SF is the place for you. Secure your exhibitor package here before we sell out!


Source: Tech Crunch

Microsoft backpedals on VR promise

If you were still waiting patiently for the virtual reality features that Microsoft promised in 2016, then I have some bad news for you. During E3 last week, Microsoft’s chief marketing officer for gaming, Mike Nichols, told GamesIndustry.biz that the company had no plans to fulfill that promise.

“We don’t have any plans specific to Xbox consoles in virtual reality or mixed reality,” Nichols told GamesIndustry.biz.

This goes against a promise that Microsoft made two years ago when Xbox chief Phil Spencer told the Verge that the Xbox One X (then dramatically known as Xbox Scorpio) would support “[the kind of] high-end VR that you see happening in the PC space.”

The release of the Xbox One X came and went without any news of VR integration, but in the interim, Microsoft did make strides toward VR and mixed reality tech for PC gaming with the release of the Windows Mixed Reality headsets for for Windows 10.

According to Nichols, it seems like Microsoft may be sticking to this PC gaming territory for awhile.

“PC is probably the best platform for more immersive VR and MR … but as it relates to Xbox, no,” he said.


Source: Tech Crunch

Species-identifying AI gets a boost from images snapped by citizen naturalists

Someday we’ll have an app that you can point at a weird bug or unfamiliar fern and have it spit out the genus and species. But right now computer vision systems just aren’t up to the task. To help things along, researchers have assembled hundreds of thousands of images taken by regular folks of critters in real life situations — and by studying these, our AI helpers may be able to get a handle on biodiversity.

Many computer vision algorithms have been trained on one of several large sets of images, which may have everything from people to household objects to fruits and vegetables in them. That’s great for learning a little about a lot of things, but what if you want to go deep on a specific subject or type of image? You need a special set of lots of that kind of image.

For some specialties, we have that already: FaceNet, for instance, is the standard set for learning how to recognize or replicate faces. But while computers may have trouble recognizing faces, we rarely do — while on the other hand, I can never remember the name of the birds that land on my feeder in the spring.

Fortunately, I’m not the only one with this problem, and for years the community of the iNaturalist app has been collecting pictures of common and uncommon animals for identification. And it turns out that these images are the perfect way to teach a system how to recognize plants and animals in the wild.

Could you tell the difference?

You might think that a computer could learn all it needs to from biology textbooks, field guides, and National Geographic. But when you or I take a picture of a sea lion, it looks a lot different from a professional shot: the background is different, the angle isn’t perfect, the focus is probably off, and there may even be other animals in the shot. Even a good computer vision algorithm might not see much in common between the two.

The photos taken through the iNaturalist app, however, are all of the amateur type — yet they have also been validated and identified by professionals who, far better than any computer, can recognize a species even when it’s occluded, poorly lit, or blurry.

The researchers, from Caltech, Google, Cornell, and iNaturalist itself, put together a limited subset of the more than 1.6 million images in the app’s databases, presented this week at CVPR in Salt Lake City. They decided that in order for the set to be robust, it should have lots of different angles and situations, so they searched for species that have had at least 20 different people spot them.

The resulting set of images (PDF) still has over 859,000 pictures of over 5,000 species. These they had people annotate by drawing boxes around the critter in the picture, so the computer would know what to pay attention to. A set of images was set aside for training the system, another set for testing it.

Examples of bounding boxes being put on images.

Ironically, they can tell it’s a good set because existing image recognition engines perform so poorly on it, not even reaching 70 percent first-guess accuracy. The very qualities that make the images themselves so amateurish and difficult to parse make them extremely valuable as raw data; these pictures haven’t been sanitized or set up to make it any easier for the algorithms to sort through.

Even the systems created by the researchers with the iNat2017 set didn’t fare so well. But that’s okay — finding where there’s room to improve is part of defining the problem space.

The set is expanding, as others like it do, and the researchers note that the number of species with 20 independent observations has more than doubled since they started working on the dataset. That means iNat2018, already under development, will be much larger and will likely lead to more robust recognition systems.

The team says they’re working on adding more attributes to the set so that a system will be able to report not just species, but sex, life stage, habitat notes, and other metadata. And if it fails to nail down the species, it could in the future at least make a guess at the genus or whatever taxonomic rank it’s confident about — e.g. it may not be able to tell if it’s anthopleura elegantissima or anthopleura xanthogrammica, but it’s definitely an anemone.

This is just one of many parallel efforts to improve the state of computer vision in natural environments; you can learn more about the ongoing collection and competition that leads to the iNat datasets here, and other more class-specific challenges are listed here.


Source: Tech Crunch

A huge spreadsheet naming ICE employees gets yanked from GitHub and Medium

A massive database of current U.S. Immigration and Customs Enforcement (ICE) employees scraped from public LinkedIn profiles has been removed from the tech platforms hosting the data. The project was undertaken by Sam Lavigne, self-described artist, programmer and researcher in response to recent revelations around ICE’s detention practices at the southern U.S. border.

Lavigne posted the database to GitHub on Tuesday and by Wednesday the repository had been removed. The database included the name, profile photo, title and city area of every ICE employee who listed the agency as their employer on the professional networking site. A more in-depth version of the data pulled all public LinkedIn data from the pool of users, including previous employment, education history and any other information those users opted to make public. The total database lists this information for 1,595 ICE employees, from the company’s CTO on down to low-level workers.

The project accompanied a Medium post about the project’s aims that has since been removed by the platform:

While I don’t have a precise idea of what should be done with this data set, I leave it here with the hope that researchers, journalists and activists will find it useful…

I find it helpful to remember that as much as internet companies use data to spy on and exploit their users, we can at times reverse the story, and leverage those very same online platforms as a means to investigate or even undermine entrenched power structures. It’s a strange side effect of our reliance on private companies and semi-public platforms to mediate nearly all aspects of our lives.

The data set appears to have violated GitHub and Medium guidelines against doxing. Medium’s anti-harassment policy specifically forbids doxing and defines it broadly, preventing “the aggregation of publicly available information to target, shame, blackmail, harass, intimidate, threaten, or endanger.”

Because it doesn’t include personal identifying information like home addresses, phone numbers or other non-public details, Lavigne’s project isn’t really doxing in the normal sense of the word, though that hasn’t made it less controversial.

GitHub’s own policy leading to the data’s removal is less clear, though the company told The Verge the repository was removed due to “doxxing and harassment.” The platform’s terms of service forbid uses of GitHub that “violate the privacy of any third party, such as by posting another person’s personal information without consent.” This leaves some room for interpretation, and it is not clear that data from a public-facing social media profile is “personal” under this definition. GitHub allows researchers to scrape data from external sites in order to aggregate it “only if any publications resulting from that research are open access.”

While Lavigne’s aggregation efforts were deemed off-limits by some tech platforms, they do raise compelling questions. What kinds of public data, in aggregate, run afoul of anti-harassment rules? Why can this kind of data be scraped for the purposes of targeted advertising or surveillance by law enforcement but not be collected in a user-facing way? The ICE database raised these questions and plenty more, but for some tech companies the question of hosting the data proved too provocative from the start.


Source: Tech Crunch

Happn takes on Tinder Places with an interactive map of missed connections

Dating app Happn, whose “missed connections” type of dating experience connects people who have crossed paths in real life, is fighting back at Tinder. Seemingly inspired by Happn’s location-based features, Tinder recently began piloting something called Tinder Places – a feature that tracks your location to match you with those people who visit your same haunts – like a favorite bar, bookshop, gym, restaurant, and more.

Of course Tinder’s move into location-based dating should worry Happn, which had built its entire dating app around the idea of matching up people who could have met in real life, but just missed doing so.

Now, Happn is challenging Tinder Places with a new feature of its own. It’s debuting an interactive map where users can discover those people they’ve crossed paths with over the past seven days.

Happn founder, French entrepreneur Didier Rappaport, dismisses the Tinder threat.

“We don’t see it as a threat at all but as a good thing,” he tells TechCrunch. “Find the people you’ve crossed paths with has always been in Happn’s DNA since the beginning….We are very flattered that Tinder wants to include the same feature in its product. However, we will never use the swipe in our product,” he says.

Rappaport believes swiping is wrong because it makes you think of the other person as a product, and that’s not Happn’s philosophy.

“We want to [give our users a chance] to interact or not with a person, to take their time to decide, to be able to move back in their timeline if suddenly they change their mind and want to have a second chance,” he notes.

To use Happn’s map, you’ll tap on a specific location you’ve visited, and are then presented with potential matches who have been there too, or within 250 meters of that spot. The map will use the same geolocation data that Happn already uses to create its timeline, but just displays it in another form.

For those who aren’t comfortable sharing their location all the time with a dating app (um, everyone?), Happn also offers an “invisibility” mode that lets people hide their location during particular parts of the day – for example, while they’re at work.

While Happn’s new feature is a nice upgrade for regular users, Tinder’s location-based features – we’re sorry to report – are more elegantly designed.

Today, Happn’s invisibility mode has to be turned on when you want to use it, or you have to pay for a subscription to schedule to come on automatically at certain times. That means it requires more effort to use on a day-to-day basis.

Meanwhile, Tinder Places lets you block a regular place you visit – like, say, the gym – from ever being recorded as a place you want to show up for matches. It also automatically removes places that would be inappropriate, including your home and work addresses, and alerts you when it’s adding a new one – so you can quickly take action to remove it, if you choose. Tinder Places is also free. (It’s just not rolled out worldwide at this time).

Happn, however, does offer a way to hide your profile information and other details from select users, and never shows your current location in real time, also like Tinder.

Happn, which launched back in 2014, now claims nearly 50 million users worldwide, across 50 major cities and 40 countries. It claims to have 6.5 million monthly users – but that’s much smaller, compared with Tinder’s estimated 50 million actives.

And with Tinder parent Match Group snatching up Hinge, suing Bumble, and effectively copying the idea of using “missed connections,” one has to wonder how much life rival dating apps, especially those of Happn’s size, have left.

The app is a free download on the App Store, Play Store and Windows Store.


Source: Tech Crunch

Google StreetView cars to help map pollution in London

From next month two Google StreetView cars will be driving around London’s streets fitted with sensors that take air quality readings every 30 meters to map and monitor air quality in the UK capital.

There will also be 100 fixed sensors fitted to lampposts and buildings in pollution blackspots and sensitive locations in the city — creating a new air quality monitoring network that Sadiq Khan, London’s mayor, is billing as “the most sophisticated in the world”.

The goal with the year-long project is to generate hyperlocal data to help feed policy responses. Khan has made tackling air pollution one of his priorities.

It’s not the first time StreetView cars have been used as a vehicle for pollution monitoring. Three years ago sensors made by San Francisco startup Aclima were fitted to the cars to map air quality in the Bay Area.

The London project is using sensors made by UK company Air Monitors.

The air quality monitoring project is a partnership between the Greater London Authority and C40 Cities network — a coalition of major cities around the world which is focused on tackling climate change and increasing health and well-being.

The project is being led by the charity Environmental Defense Fund Europe, in partnership with Air Monitors, Google Earth Outreach, Cambridge Environmental Research Consultants, University of Cambridge, National Physical Laboratory, and the Environmental Defense Fund team in the United States.

King’s College London will also be undertaking a linked study focused on schools.

Results will be shared with members of the C40 Cities network — with the ambition of developing policy responses that help improve air quality for hundreds of millions of city dwellers around the world.


Source: Tech Crunch

Trump signs an executive order to detain families together at the border indefinitely

President Trump has signed an executive order to reverse a practice recently enacted by his own administration that resulted in the separation of children from their families at the border.

The language of the executive order, titled “Affording Congress an Opportunity to Address Family Separation,” points blame at Congress, echoing Trump’s previous statements demanding that this issue be resolved through legislation although it was not implemented through legislation.

The meat of the order:

“Section 1. Policy. It is the policy of this Administration to rigorously enforce our immigration laws. Under our laws, the only legal way for an alien to enter this country is at a designated port of entry at an appropriate time. When an alien enters or attempts to enter the country anywhere else, that alien has committed at least the crime of improper entry and is subject to a fine or imprisonment under section 1325(a) of title 8, United States Code. This Administration will initiate proceedings to enforce this and other criminal provisions of the INA until and unless Congress directs otherwise. It is also the policy of this Administration to maintain family unity, including by detaining alien families together where appropriate and consistent with law and available resources. It is unfortunate that Congress’s failure to act and court orders have put the Administration in the position of separating alien families to effectively enforce the law.”

The executive order proposes a “temporary detention policy” that would allow the Department of Homeland Security to detain families attempting to enter the United States at the southern border “during the pendency of any criminal improper entry or immigration proceedings involving their members.”

That portion of the order suggests that DHS would indefinitely detain a family, together while any of its members await prosecution and potential deportation, a policy that looks likely to violate a court decision known as the Flores Agreement. Because that legal precedent forbids the indefinite detainment of children at the border, the administration is likely gearing up for a clash in the courts.

As controversy around the southern border erupted in recent days, many major tech companies weighed in with vocal opposition to the Trump administration’s recent practice of separating adults who enter the U.S. illegally from the children they bring with them. Microsoft’s Satya Nadella also denounced the policy, though the company is facing both internal and external criticism over its previously announced intentions to supply deep learning and facial recognition software to U.S. Immigration and Customs Enforcement (ICE) through a lucrative federal contract.


Source: Tech Crunch

Microsoft is buying AI startup, Bonsai

If all of the big tech co’s agree on one thing at the moment, it’s that artificial intelligence and machine learning point the way forward for their businesses. As a matter of fact, Microsoft is about to acquire Bonsai, a small Berkeley-based startup it hopes to make the centerpiece of its AI efforts.

The company specializes in reinforcement learning, a kind of trial and error approach to teach a system within in the confines of a simulation. That learning can be used train autonomous systems to complete specific tasks.  Microsoft says the acquisition will serve to forward the kind of research the company has been pursuing in the field by leveraging its Azure cloud platform.

“To realize this vision of making AI more accessible and valuable for all, we have to remove the barriers to development, empowering every developer, regardless of machine learning expertise, to be an AI developer,” Microsoft Corporate VP Gurdeep Pall said in an announcement. “Bonsai has made tremendous progress here and Microsoft remains committed to furthering this work.”

Microsoft is among a number of high profile companies that have supported the four-year-old startup. Last year, it joined ABB, Samsung and Siemens in helping the company raise a $7.6 million round, bringing the company’s total raise up to $13.6 million, per CrunchbaseThe pending  move follows the recent high-profile acquisition of code hosting tool, GitHub.

“Going forward, we see a massive opportunity to empower enterprises & developers globally with the tools and technology needed to build and operate the BRAINs that power these intelligent autonomous systems,” Bonsai co-founder/CEO Mark Hammond said in a blog post. “We are not the only ones that feel this way. Today we are excited to announce that Microsoft will be acquiring Bonsai to help accelerate the realization of this common vision.”


Source: Tech Crunch