Auteur/autrice : noflux (Page 54 of 625)

Twitter is testing emoji reactions for tweets

10 septembre 2021 / noflux

“Emoji reactions have been available for years on other services like Facebook, as well as on Twitter itself within direct messages. But what’s interesting about Twitter’s emoji choices for its latest test is that none of them are especially negative. There’s no “Angry face” like you’ll find on Facebook, or “Thumbs down” like in Twitter’s direct message emoji reactions.Twitter explains that it decided against choosing these negative emoji because people it surveyed said “they were concerned about receiving negative reactions to some of their thoughts.” A valid concern given how toxic many conversations on Twitter can be.”

Source : Twitter is testing emoji reactions for tweets – The Verge

Climate activist arrested after ProtonMail provided his IP address : ProtonMail

6 septembre 2021 / noflux

First, unlike other providers, ProtonMail does fight on behalf of users. Few people know this (it’s in our transparency report), but we actually fought over 700 cases in 2020 alone, which is a huge amount. This particular case however could not be fought.

Second, ProtonMail is one of the only email providers that provides a Tor onion site for anonymous access. This allows users to connect to ProtonMail through the Tor anonymity network. You can find more information here: protonmail.com/tor

Third, no matter what service you use, unless it is based 15 miles offshore in international waters, the company will have to comply with the law. This case does illustrate one benefit of ProtonMail’s Swiss jurisdiction, as no less than 3 authorities in 2 countries were required to approve the request, which is a much higher bar than most other jurisdictions. Under Swiss law, it is also obligatory for the suspect to be notified that their data was requested.

via Reddit

Delays Aren’t Good Enough—Apple Must Abandon Its Surveillance Plans | Electronic Frontier Foundation

6 septembre 2021 / noflux

apple with an eye in the center

“The features Apple announced a month ago, intending to help protect children, would create an infrastructure that is all too easy to redirect to greater surveillance and censorship. These features would create an enormous danger to iPhone users’ privacy and security, offering authoritarian governments a new mass surveillance system to spy on citizens. They also put already vulnerable kids at risk, especially LGBTQ youth, and create serious potential for danger to children in abusive households. The responses to Apple’s plans have been damning: over 90 organizations across the globe have urged the company not to implement them, for fear that they would lead to the censoring of protected speech, threaten the privacy and security of people around the world, and have disastrous consequences for many children.”

Source : Delays Aren’t Good Enough—Apple Must Abandon Its Surveillance Plans | Electronic Frontier Foundation

Une action en justice pourrait bloquer l’accès aux plus importants sites pornographiques

5 septembre 2021 / noflux

Un panneau Pornhub à Las Vegas, Nevada, en 2017.

“Dans les faits, il est en réalité souvent compliqué pour la justice de remonter jusqu’aux propriétaires des grands sites pornographiques, qui s’abritent derrière des cascades de sociétés domiciliées dans plusieurs pays. MindGeek, l’entreprise leader du secteur – elle possède Pornhub, RedTube ou encore YouPorn, pour un chiffre d’affaires qui se compte en centaines de millions de dollars –, a ses bureaux au Canada, mais est fiscalement domiciliée au Luxembourg et possède des dizaines de filiales dans une demi-douzaine de pays. Les associations, en coordination avec la chancellerie, ont donc changé d’angle d’attaque : plutôt que les sites, elles ont ciblé les fournisseurs d’accès Internet (FAI), présents sur le territoire, Orange, SFR, Free, Bouygues… « Nous assignons les FAI en référé car il n’est pas toujours possible d’identifier les éditeurs de contenus pornographiques », reconnaît Me Laurent Bayon, le conseil des deux associations, qui espère que le ministère public appuiera leur demande.”

Source : Une action en justice pourrait bloquer l’accès aux plus importants sites pornographiques

16 juillet 2021 / noflux

“We’ve enhanced Android’s auto-rotate feature with face detection, using the front-facing camera to more accurately recognize when to rotate the screen. This is especially helpful for people who are using their devices while lying down on a couch or in bed, for example. For developers, this means that the auto-rotation behavior will provide a better user experience for users who have opted in through Settings. The enhanced auto-rotate feature lives within our recently announced Private Compute Core, so images are never stored or sent off the device. In Beta 3 this feature is available on Pixel 4 and later Pixel devices.To make screen rotation as speedy as possible on all devices, we’ve also optimized the animation and redrawing and added an ML-driven gesture-detection algorithm. As a result, the latency for the base auto-rotate feature has been reduced by 25%, and the benefits of the face detection enhancement build on top of those improvements. Give the auto-rotate improvements a try and let us know what yo”

Source : Android Developers Blog: Android 12 Beta 3 and final APIs

default filename tv 📺

4 juillet 2021 / noflux

“default filename tv finds and plays youtube videos that were uploaded from the camera without edits to the filename.”

Source : default filename tv 📺

25 juin 2021 / noflux

“Harness the power of AI to quickly turn simple brushstrokes into realistic landscape images for backgrounds, concept exploration, or creative inspiration. 🖌️ The NVIDIA Canvas app lets you create as quickly as you can imagine.”

via NVIDIA Studio

Reverse engineering generative models from a single deepfake image

22 juin 2021 / noflux

“Deepfakes have become more believable in recent years. In some cases, humans can no longer easily tell some of them apart from genuine images. Although detecting deepfakes remains a compelling challenge, their increasing sophistication opens up more potential lines of inquiry, such as: What happens when deepfakes are produced not just for amusement and awe, but for malicious intent on a grand scale? Today, we — in partnership with Michigan State University (MSU) — are presenting a research method of detecting and attributing deepfakes that relies on reverse engineering from a single AI-generated image to the generative model used to produce it. Our method will facilitate deepfake detection and tracing in real-world settings, where the deepfake image itself is often the only information detectors have to work with.”

Source : Reverse engineering generative models from a single deepfake image

AI Explorables | PAIR

21 juin 2021 / noflux

“The rapidly increasing usage of machine learning raises complicated questions: How can we tell if models are fair? Why do models make the predictions that they do? What are the privacy implications of feeding enormous amounts of data into models? This ongoing series of interactive, formula-free essays will walk you through these important concepts.”

Source : AI Explorables | PAIR

LaMDA: our breakthrough conversation technology

20 juin 2021 / noflux

An animation demonstrating how language is processed by LaMDA technology.

“These early results are encouraging, and we look forward to sharing more soon, but sensibleness and specificity aren’t the only qualities we’re looking for in models like LaMDA. We’re also exploring dimensions like “interestingness,” by assessing whether responses are insightful, unexpected or witty. Being Google, we also care a lot about factuality (that is, whether LaMDA sticks to facts, something language models often struggle with), and are investigating ways to ensure LaMDA’s responses aren’t just compelling but correct. But the most important question we ask ourselves when it comes to our technologies is whether they adhere to our AI Principles. Language might be one of humanity’s greatest tools, but like all tools it can be misused. Models trained on language can propagate that misuse — for instance, by internalizing biases, mirroring hateful speech, or replicating misleading information. And even when the language it’s trained on is carefully vetted, the model itself can still be put to ill use. ”

Source : LaMDA: our breakthrough conversation technology