Connect with us

Tech News

Facebook removed 14 million pieces of terrorist content this year, and the numbers are rising – TechCrunch

Published

on

Facebook must exert constant vigilance to prevent its platform from being taken over by ne’er-do-wells, but how exactly it does that is only really known to itself. Today, however, the company has graced us with a bit of data on what tools it’s using and what results they’re getting — for instance, more than 14 million pieces of “terrorist content” removed this year so far.

More than half of that 14 million was old content posted before 2018, some of which had been sitting around for years. But as Facebook points out, that content may very well have also been unviewed that whole time. It’s hard to imagine a terrorist recruitment post going unreported for 970 days (the median age for content in Q1) if it was seeing any kind of traffic.

Perhaps more importantly, the numbers of newer content removed (with, to Facebook’s credit, a quickly shrinking delay) appear to be growing steadily. In Q1, 1.2 million items were removed; in Q2, 2.2 million; in Q3, 2.3 million. User-reported content removals are growing as well, though they are much smaller in number — around 16,000 in Q3. 99 percent of it, Facebook proudly reports, is removed “proactively.”

Something worth noting: Facebook is careful to avoid positive or additive verbs when talking about this content, for instance it won’t say that “terrorists posted 2.3 million pieces of content,” but rather that was the number of “takedowns” or content “surfaced.” This type of phrasing is more conservative and technically correct, as they can really only be sure of their own actions, but it also serves to soften the fact that terrorists are posting hundreds of thousands of items monthly.

The numbers are hard to contexualize. Is this a lot or a little? Both, really. The amount of content posted to Facebook is so vast that almost any number looks small next to it, even a scary one like 14 million pieces of terrorist propaganda.

It is impressive, however, to hear that Facebook has greatly expanded the scope of its automated detection tools:

Our experiments to algorithmically identify violating text posts (what we refer to as “language understanding”) now work across 19 languages.

And it fixed a bug that was massively slowing down content removal:

In Q2 2018, the median time on platform for newly uploaded content surfaced with our standard tools was about 14 hours, a significant increase from Q1 2018, when the median time was less than 1 minute. The increase was prompted by multiple factors, including fixing a bug that prevented us from removing some content that violated our policies, and rolling out new detection and enforcement systems.

The Q3 number is two minutes. It’s a work in progress.

No doubt we all wish the company had applied this level of rigor somewhat earlier, but it’s good to know that the work is being done. Notable is that a great deal of this machinery is not focused on simply removing content, but on putting it in front of the constantly growing moderation team. So the most important bit is still, thankfully and heroically, done by people.

Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tech News

Fujifilm GFX100 can now produce 400MP images with a firmware update

Published

on

Given the onslaught of smartphones, camera makers have had to step up to offer features that would be difficult to pull off with a mobile device. Most have started turning DSLRs and ILCs into webcams, for example, to ride on the wave of video conferencing platforms today. Fujifilm has yet another such feature that would probably be impossible with smartphones today, allowing the 102MP GFX100 mirrorless camera to generate pictures with 400MP resolutions.

It almost sounds like the opposite of pixel binning, where smartphones use adjacent pixels to produce an image that has a smaller resolution but a higher level of detail. The principle behind Fujifilm’s “Pixel Shift Multi-Shot” however, isn’t exactly new and can be found in cameras and smartphones under different names. In a nutshell, it takes multiple shots at slightly shifted positions which are then combined into a single 400MP image.

When the user presses the shutter button, the GFX100 records 16 frames in rapid succession, using the in-body image stabilization or IBIS to shift the sensor slightly at each frame. It doesn’t end in the camera, however, and requires a separate Pixel Shift Combiner software for PCs and Macs to combine those 16 frames into a single RAW file in DNG format. A different piece of software like Capture One can then turn that into a 400MP TIFF or JPEG.

The resulting image naturally contains more detail than an equivalent 100MP photo. Fujifilm claims that these 400MP images are able to reproduce more faithfully the details and colors of their subjects. Of course, that also means that file sizes are exponentially larger, too, with a single JPEG taking up hundreds of megabytes.

That might be a deal-breaker for hobbyists and even some pros but Fujifilm might not have those users in mind anyway. It is positioning this feature for digitally archiving artworks, architectures, and other cultural assets where storage is a small price to pay for being able to faithfully preserve such artifacts. Pixel Shift Multi-Shot is available in the latest firmware update to the Fujifilm GFX100 but the Pixel Shift Combiner and Capture One programs have to be downloaded separately.

Continue Reading

Tech News

TikTok given another seven days to sell its US operations

Published

on

Almost like some of the smooth dance moves you can see on the social network, TikTok has again narrowly dodged what could be the fatal deadline that would end it, at least in the US. As the new November 27 deadline approaches, owner ByteDance was given another week to convince the US government not to ban it outright. That remains a quest of Herculean proportions as the Chinese company’s multiple proposals have been repeatedly rejected already.

TikTok was supposed to go dark in the US on November 12, the initial deadline set by US President Donald Trump for the company to come up with a proposal to sell its US operations to a US owner. Whether it was because of the elections or not, that date approached with no further word from the government, leading TikTok to wonder if they were miraculously forgotten. It wasn’t but it at least got a 15-day reprieve.

Those fifteen days will be over on Friday and yet no resolution has been reached. It seems, however, that the Trump administration is willing to draw the uncertainty out as long as it can and filed for another seven days for ByteDance to find a buyer. The problem is that it seems that the US government is still not satisfied with the terms.

For quite a few months already, TikTok’s owner has been in talks with Walmart, Oracle, and other US investors to create an entity that will handle the US side of TikTok’s operations, including the critical data of American users. ByteDance revealed it already made four other proposals before November 10 and an insider tip claims it submitted yet another one. The extension is supposedly meant to give the US Treasury Department time to review that proposal.

This, of course, leaves TikTok’s fate still hanging in the balance, especially given the current uncertainty in the US political scene. The Trump administration itself, however, is also running out of time before it hands over the reins to a new government. Previous attempts to indirectly curtail TikTok’s operations were already shot down in federal courts, leaving this divestment strategy as its last recourse.

Continue Reading

Tech News

Windows 10 might install and run Android apps directly next year

Published

on

Once upon a time, Windows was the operating system that ran the world, with Mac OS a distant second and Linux holding an invisible world of networks and servers. Today, however, not only have Linux and macOS risen up the ranks, there is also Android, iOS, iPadOS, and even Chrome OS to contend with. Rather than completely conceding defeat, Microsoft is instead embracing some of these rival platforms, with rumors of a deeper integration with Android that would let its mobile apps run directly on Windows 10 by 2021 at the earliest.

It is, of course, already possible to run Android apps on Windows. There are emulators and tools like Bluestack that have focused more on games as well as Microsoft’s own Your Phone utility that can run apps from a compatible Samsung Galaxy phone as if they were native. Based on Windows Central’s brief blurb, Microsoft may be aiming for something even more seamless that would let users install Android apps from its Microsoft Store.

To some extent, this would mirror Chrome OS’s own ability to run Android apps somewhat natively, which requires supporting the Android runtime itself on Windows. Given Android’s open nature and Microsoft’s own expertise, not to mention its work with the Windows Subsystem for Linux, it’s not entirely out of the realm of possibility. Running Android apps natively, however, is just one part of the equation and Microsoft will have to do a lot more work to make it a convincing and attractive feature.

The biggest question mark will be Google Play support, something that Google will probably be not so keen on allowing. Without Google Play Store and services, Microsoft will either have to follow Huawei or, at least, set up its own app store ala Samsung. Neither, however, would be an ideal situation.

The report also has some news for Windows 10X which might end up DOA by the time it launches in Spring 2021. Not only will this “lightweight” edition of Windows 10 run only on traditional laptops and tablets, it won’t be getting support for win32 applications until 2022. Microsoft will allegedly rely more on “Cloud PC” app streaming and perhaps support for Android apps to make do until then.

Continue Reading

Trending