Kenya's National Cohesion and Integration Commission (NCIC), a government agency that aims to eradicate ethnic or racial discrimination among the country's 45 tribes, has given Facebook seven days to tackle hate speech related to next month's election on its platform. If the social media fails to do so, it faces suspension in the country. The agency's warning comes shortly after international NGO Global Witness and legal non-profit Foxglove released a report detailing how Facebook approved ads written to instigate ethnic violence in both English and Swahili.

The organizations joined forces to conduct a study testing Facebook's ability to detect hate speech and calls for ethnic-based violence ahead of the Kenyan elections. As Global Witness explained in its report, the country's politics are polarized and ethnically driven — after the 2007 elections, for instance, 1,300 people were killed and hundreds of thousands more had to flee their homes. A lot more people use social media today compared to 2007, and over 20 percent of the Kenyan population is on Facebook, where hate speech and misinformation are major issues.

The groups decided not to publish the exact ads they submitted for the test because they were highly offensive, but they used real-life examples of hate speech commonly used in Kenya. They include comparisons of specific tribal groups to animals and calls for their members' rape, slaughter and beheading. "Much to our surprise and concern," Global Witness reported, "all hate speech examples in both [English and Swahili] were approved." The NCIC said the NGOs' report corroborates its own findings. 

After the organizations asked Facebook for a comment regarding what it had discovered and hence made it aware of the study, Meta published a post that details how it is preparing for Kenya's election. In it, the company said it has built a more advanced content detection technology and has hired dedicated teams of Swahili speakers to help it "remove harmful content quickly and at scale." To see if Facebook truly has implemented changes that has improved its detection system, the organizations resubmitted its test ads. They were approved yet again. 

In a statement sent to both Global Witness and Gizmodo, Meta said it has taken "extensive steps" to "catch hate speech and inflammatory content in Kenya" and that the company is "intensifying these efforts ahead of the election." It also said, however, that there will be instances where it misses things " as both machines and people make mistakes."

Global Witness said its study's findings follow a similar pattern it previously uncovered in Myanmar, where Facebook played a role in enabling calls for ethnic cleansing against Rohingya Muslims. It also follows a similar pattern the organization unearthed in Ethiopia wherein bad actors used the Facebook to incite violence. The organizations and Facebook whistleblower Frances Haugen are now calling on Facebook to implement the "Break the Glass” package of emergency measures it took after the January 6th, 2021 attack on the US Capitol. They's also asking the social network to suspend paid digital advertisements in Kenya until the end of the elections on August 9th. 

TikTok is conducting a broader test of games in its all-conquering app. The company recently added a way for creators in some markets (including the US) to append one of nine mini-games to a video by tapping the Add Link button and choosing the MiniGame option. When viewers come across a video that links to a game, they can start playing it by tapping a link next to the creator's username.

“Currently, we’re exploring bringing HTML5 games to TikTok through integrations with third-party game developers and studios," a TikTok spokesperson told TechCrunch. One of the games is from Aim Lab, the maker of a popular aim training app of the same name. Its TikTok game is called Mr. Aim Lab’s Nightmare. TikTok's other partners on the initiative include developers Voodoo, Nitro Games, FRVR and Lotem.

None of the games have ads or in-app purchases at the minute and the project is in the early stages of testing. TikTok is looking to find out how (or if) creators craft content around them, and how users interact with the games. As The Verge notes, users can record their gameplay and share it in a fresh video.

Reports in recent months suggested TikTok was readying for a major push into gaming. Parent company ByteDance bought game developer Moonton Technology last year. TikTok teamed up with Zynga for an exclusive mobile game called Disco Loco 3D; a charity game called Garden of Good, through which players can trigger donations to Feeding America, became available on the US version of TikTok in June. TikTok previously tested HTML5 games in Vietnam.

Other major tech companies have made a push into mobile gaming, including Apple, Google and, more recently, Netflix. Zynga, of course, became a social gaming giant with the help of Facebook's massive reach, while Facebook moved into cloud gaming in 2020. It's no secret that Meta is trying to ape many of TikTok's features across many of its apps, so it's interesting to see TikTok taking a leaf out of Facebook's playbook on the gaming front.

You can find AI that creates new images, but what if you want to fix an old family photo? You might have a no-charge option. Louis Bouchard and PetaPixel have drawn attention to a free tool recently developed by Tencent researchers, GFP-GAN (Generative Facial Prior-Generative Adversarial Network), that can restore damaged and low-resolution portraits. The technology merges info from two AI models to fill in a photo's missing details with realistic detail in a few seconds, all the while maintaining high accuracy and quality.

Conventional methods fine-tune an existing AI model to restore images by gauging differences between the artificial and real photos. That frequently leads to low-quality results, the scientists said. The new approach uses a pre-trained version of an existing model (NVIDIA's StyleGAN-2) to inform the team's own model at multiple stages during the image generation process. The technique aims to preserve the "identity" of people in a photo, with a particular focus on facial features like eyes and mouths.

You can try a demo of GFP-GAN for free. The creators have also posted their code to let anyone implement the restoration tech in their own projects.

This project is still bound by the limitations of current AI. While it's surprisingly accurate, it's making educated guesses about missing content. The researchers warned that you might see a "slight change of identity" and a lower resolution than you might like. Don't rely on this to print a poster-sized photo of your grandparents, folks. All the same, the work here is promising — it hints at a future where you can easily rescue images that would otherwise be lost to the ravages of time.

In late 2020, Alphabet's DeepMind division unveiled its novel protein fold prediction algorithm, AlphaFold, and helped solve a scientific quandary that had stumped researchers for half a century. In the year since its beta release, half a million scientists from around the world have accessed the AI system's results and cited them in their own studies more than 4,000 times. On Thursday, DeepMind announced that it is increasing that access even further by radically expanding its publicly-available AlphaFold Protein Structure Database (AlphaFoldDB) — from 1 million entries to 200 million entries.

Alphabet partnered with EMBL’s European Bioinformatics Institute (EMBL-EBI) for this undertaking, which covers proteins from across the kingdoms of life — animal, plant, fungi, bacteria and others. The results can be viewed on the UniProt, Ensembl, and OpenTargets websites or downloaded individually via GitHub, "for the human proteome and for the proteomes of 47 other key organisms important in research and global health," per the AlphaFold website.

"AlphaFold is the singular and momentous advance in life science that demonstrates the power of AI," Eric Topol, Founder and Director of the Scripps Research Translational Institute, siad in a press statement Thursday. "Determining the 3D structure of a protein used to take many months or years, it now takes seconds. AlphaFold has already accelerated and enabled massive discoveries, including cracking the structure of the nuclear pore complex. And with this new addition of structures illuminating nearly the entire protein universe, we can expect more biological mysteries to be solved each day."

AlphaFold has been leveraged in a number of applications, from advancing Leprosy and Chagas Disease research to conserving honeybees and addressing plastic pollution. DeepMind has also developed AIs that can outmatch top human players, master games without even knowing the rules and even improve traffic patterns. DeepMind's co-founder, Mustafa Suleyman, left the company in January in order to launch a new venture, Inflection.AI, with LinkedIn co-founder Reid Hoffman, shortly before accusations of ongoing sexual harassment were leveled at the company by a former employee.

Tim Hortons has agreed to settle multiple class action lawsuits that accused the company of tracking customers' locations through its app without consent. Under the proposed settlement, which requires a judge's approval, eligible customers in Canada will receive a free hot drink and baked good. In other words, in exchange for your location data, the restaurant chain will give you coffee and a doughnut.

In an email to customers, the company said it will delete any geolocation data it obtained from them between April 1st, 2019 and September 30th, 2020, and tell third-party vendor Radar Labs to do the same. An investigation conducted by Canadian privacy officials determined last month that the Tim Hortons app was tracking and recording users' locations every few minutes, even when they didn't have the app open. The probe determined that the company and Radar Labs didn't have sufficient consent from users for that level of tracking. 

Tim Hortons has not admitted to any wrongdoing and isn't believed to have misused the data. It also avoided disciplinary action.

If you’ve been patiently waiting for Valve to ship that Steam Deck you reserved months ago, you may not have to wait too much longer to get your hands on the device. The company claims it will be able to fulfill all current Steam Deck reservations by the end of the year.

Valve says that many of its supply chain issues are easing and it’s able to continue increasing production. It has moved some people’s scheduled deliveries up to the third quarter (i.e. between now and the end of September). All other outstanding reservations are now scheduled for delivery in Q4. People who lock in a reservation right now should still be able to get one in time for the holiday season. After Valve reaches its capacity for that period, it will start filling up the delivery queue for early 2023. 

The update comes a month after Valve said it was doubling the number of weekly Steam Deck shipments. This is a rare sliver of positive news for a sought-after piece of gaming hardware, given the supply chain problems that have stymied PlayStation 5 and Xbox Series X production. Who knows? We might even see the Steam Deck on retail shelves before too long.

Blue Origin has revealed when New Shepard's sixth crewed flight will take place. The NS-22 mission (New Shepard's 22nd launch overall) is set for August 4th, which is two months to the day after its previous spaceflight. The launch window opens at 9:30AM ET.

The company recently announced the passengers who will travel to the edge of space this time around. Among them are engineer Sara Sabry and entrepreneur Mário Ferreira, who will be the first Egyptian and Portuguese people to visit space, respectively. Mountaineer Vanessa O’Brien, meanwhile, will become the first woman to reach the summit of Mount Everest, venture to Challenger Deep (believed to be the deepest point of the ocean) and cross the Kármán line — the boundary between Earth's atmosphere and outer space.

The crew also includes Coby Cotton, a cofounder of the popular trickshot and comedy YouTube channel Dude Perfect, and telecoms executive Steve Young. The sixth passenger is Clint Kelly III, who started the Autonomous Land Vehicle project at the Defense Advanced Research Projects Agency (DARPA) in the 1980s. Blue Origin notes that Kelly is credited with creating the technology base that paved the way for modern self-driving vehicles.

Blue Origin

Twitter has published its 20th transparency report, and the details still aren't reassuring to those concerned about abuses of personal info. The social network saw "record highs" in the number of account data requests during the July-December 2021 reporting period, with 47,572 legal demands on 198,931 accounts. The media in particular faced much more pressure. Government demands for data from verified news outlets and journalists surged 103 percent compared to the last report, with 349 accounts under scrutiny.

The largest slice of requests targeting the news industry came from India (114), followed by Turkey (78) and Russia (55). Governments succeeded in withholding 17 tweets.

As in the past, US demands represented a disproportionately large chunk of the overall volume. The country accounted for 20 percent of all worldwide account info requests, and those requests covered 39 percent of all specified accounts. Russia is still the second-largest requester with 18 percent of volume, even if its demands dipped 20 percent during the six-month timeframe.

The company said it was still denying or limiting access to info when possible. It denied 31 percent of US data requests, and either narrowed or shut down 60 percent of global demands. Twitter also opposed 29 civil attempts to identify anonymous US users, citing First Amendment reasons. It sued in two of those cases, and has so far had success with one of those suits. There hasn't been much success in reporting on national security-related requests in the US, however, and Twitter is still hoping to win an appeal that would let it share more details.

Calls for data spiked starting in early 2020, and this latest transparency report indicates that they're only continuing to climb. Twitter sees the attempts to target journalists as a mounting threat to freedom of the press. We'd add that the situation is volatile, at least in the near term. Russia cut off access to Twitter following that country's invasion of Ukraine, and the company is suing India to resist an order to block accounts. Don't be surprised if the next report paints a significantly different picture.

Twitter has been testing out a few experimental features lately, including tweets that can be co-authored by two accounts and an AIM- or MySpace-style status label. For its latest trick, Twitter is toying around with a way for people to pack a single tweet with multiple images, videos and GIFs.

“We’re testing a new feature with select accounts for a limited time that will allow people to mix up to four media assets into a single tweet, regardless of format. We’re seeing people have more visual conversations on Twitter and are using images, GIFs and videos to make these conversations more exciting," Twitter told TechCrunch in a statement. "With this test we’re hoping to learn how people combine these different media formats to express themselves more creatively on Twitter beyond 280 characters.”

Although we haven't seen these tweets in the wild as yet, app researcher Alessandro Paluzzi previously shared screenshots of what the tweet composer looks like when adding several forms of media. It looks similar to the existing method of attaching multiple images to a tweet, with the option to add and remove photos, videos and GIFs and to shuffle the order of them around. Having multiple videos or GIFs in a single tweet could end up looking messy, though, and it might cause havoc on people's data plans.

Framework sold its eponymous laptop on the promise that end users should feel comfortable enough to fix almost any hardware problem themselves. Replacing a component shouldn’t be the reserve of dedicated service professionals if all you need is a T5 screwdriver and patience.

When the company released its new 12th-generation Intel Core mainboards, it couriered over a new board which could be inserted into last year’s model. And given that I don’t consider myself to be a very confident DIY-er, it made sense for me to put Framework’s promises to the test.

As you can see in the video below, laptops aren’t yet at the stage where you can pull components out as if they were Lego bricks. Although I think the industry is missing a trick by not making these components a lot simpler to assemble by standardizing the connections.

That said, one of the biggest hurdles was the ZIF connectors, which briefly made me wonder if I was really cut out for tech journalism. Those little lay-flat ribbons may be great for space but they’re a nightmare if you’ve got big hands and poor eyesight.

But, as I said when the new hardware was released, it’s empowering just how easy this stuff can be, more or less, if companies make even the smallest bit of effort. I was able to do this, talk to the camera (which probably slowed me down by quite a bit) and not blow myself up. And if I can do this, then surely you can too.