Month: May 2024

Enlarge (credit: Boris Zhitkov | Moment)

As the US moves toward criminalizing deepfakes—deceptive AI-generated audio, images, and videos that are increasingly hard to discern from authentic content online—tech companies have rushed to roll out tools to help everyone better detect AI content.

But efforts so far have been imperfect, and experts fear that social media platforms may not be ready to handle the ensuing AI chaos during major global elections in 2024—despite tech giants committing to making tools specifically to combat AI-fueled election disinformation. The best AI detection remains observant humans, who, by paying close attention to deepfakes, can pick up on flaws like AI-generated people with extra fingers or AI voices that speak without pausing for a breath.

Among the splashiest tools announced this week, OpenAI shared details today about a new AI image detection classifier that it claims can detect about 98 percent of AI outputs from its own sophisticated image generator, DALL-E 3. It also "currently flags approximately 5 to 10 percent of images generated by other AI models," OpenAI's blog said.

Enlarge / Raspberry Pi Connect looks like a good reason to make a Pi account, at least if you're not running your own DynDNS, VPN, and other remote access schemes. (credit: Raspberry Pi)

One Raspberry Pi often leads to another. Soon enough, you're running out of spots in your free RealVNC account for your tiny boards and "real" computers. Even if you go the hardened route of SSH or an X connection, you have to keep track of where they all are. All of this is not the easiest thing to tackle if you're new to single-board computers or just eager to get started.

Enter Raspberry Pi Connect, a new built-in way to access a Rasbperry Pi from nearly anywhere you can open a browser, whether to control yourself or provide remote assistance. On a Raspberry Pi 4, 5, or Pi 400 kit, you install Pi connect with a single terminal line, reboot the Pi, and then click a new tray icon to connect the Pi to a Raspberry Pi ID (and then enable two-factor authentication, of course).

From then on, visiting connect.raspberrypi.com gives you an encrypted connection to your desktop. It's a direct connection if possible, and if not, it runs through relay servers in London, encrypting it with DTLS and keeping only the metadata needed for the service to work. The Pi will show a notification in its tray that somebody has connected, and you can manage screen sharing from there. The Pi's docs site has a lot more on the particulars.

Enlarge / Dispensers for Charged Lemondade, a caffeinated lemonade drink, at Panera Bread, Walnut Creek, California, March 27, 2023. (credit: Getty | Smith Collection/Gado)

Panera Bread will stop selling its highly caffeinated "Charged" drinks, which have been the subject of at least three lawsuits and linked to at least two deaths.

It is unclear when exactly the company will pull the plug on the potent potables, but in a statement to Ars Tuesday, Panera said it was undergoing a "menu transformation" that includes an "enhanced beverage portfolio." The company plans to roll out various new drinks, including a lemonade and tea, but a spokesperson confirmed that the new flavors would not contain added caffeine as the "charged" drinks did.

The fast-casual cafe-style chain drew national attention in 2022 for the unexpectedly high caffeine levels in the drinks, which were initially offered as self-serve with free refills.

Enlarge / Apple's latest iPad Air, now in two sizes. The Magic Keyboard accessory is the same one that you use with older iPad Airs and Pros, though they can use the new Apple Pencil Pro. (credit: Andrew Cunningham)

Apple has a brand-new lineup of iPad Pro and Air models for the first time in well over a year. Most people would probably be hard-pressed to tell the new ones from the old ones just by looking at them, but after hands-on sessions with both sizes of both tablets, the small details (especially for the Pros) all add up to a noticeably refined iPad experience.

iPad Airs: Bigger is better

But let's begin with the new Airs, since there's a bit less to talk about. The 11-inch iPad Air (technically the 6th-generation model) is mostly the same as the previous-generation A14 and M1 models, design-wise, with identical physical dimensions and weight. It's still the same slim-bezel design Apple introduced with the 2018 iPad Pro, just with a 60 Hz LCD display panel and Touch ID on the power button rather than Face ID.

So when Apple says the device has been "redesigned," the company is mainly referring to the fact that the webcam is now mounted on the long edge of the tablet rather than the short edge. This makes its positioning more laptop-y when it's docked to the Magic Keyboard or some other keyboard.

Enlarge / Dmitry Yuryevich Khoroshev, aka LockBitSupp (credit: UK National Crime Agency)

Since at least 2019, a shadowy figure hiding behind several pseudonyms has publicly gloated for extorting millions of dollars from thousands of victims he and his associates had hacked. Now, for the first time, “LockBitSupp” has been unmasked by an international law enforcement team, and a $10 million bounty has been placed for his arrest.

In an indictment unsealed Tuesday, US federal prosecutors unmasked the flamboyant persona as Dmitry Yuryevich Khoroshev, a 51-year-old Russian national. Prosecutors said that during his five years at the helm of LockBit—one of the most prolific ransomware groups—Khoroshev and his subordinates have extorted $500 million from some 2,500 victims, roughly 1,800 of which were located in the US. His cut of the revenue was allegedly about $100 million.

Damage in the billions of dollars

“Beyond ransom payments and demands, LockBit attacks also severely disrupted their victims' operations, causing lost revenue and expenses associated with incident response and recovery,” federal prosecutors wrote. “With these losses included, LockBit caused damage around the world totaling billions of US dollars. Moreover, the data Khoroshev and his LockBit affiliate co-conspirators stole—containing highly sensitive organizational and personal information—remained unsecure and compromised in perpetuity, notwithstanding Khoroshev’s and his co-conspirators' false promises to the contrary.”

Enlarge (credit: Getty Images)

Microsoft has introduced a GPT-4-based generative AI model designed specifically for US intelligence agencies that operates disconnected from the Internet, according to a Bloomberg report. This reportedly marks the first time Microsoft has deployed a major language model in a secure setting, designed to allow spy agencies to analyze top-secret information without connectivity risks—and to allow secure conversations with a chatbot similar to ChatGPT and Microsoft Copilot. But it may also mislead officials if not used properly due to inherent design limitations of AI language models.

GPT-4 is a large language model (LLM) created by OpenAI that attempts to predict the most likely tokens (fragments of encoded data) in a sequence. It can be used to craft computer code and analyze information. When configured as a chatbot (like ChatGPT), GPT-4 can power AI assistants that converse in a human-like manner. Microsoft has a license to use the technology as part of a deal in exchange for large investments it has made in OpenAI.

According to the report, the new AI service (which does not yet publicly have a name) addresses a growing interest among intelligence agencies to use generative AI for processing classified data, while mitigating risks of data breaches or hacking attempts. ChatGPT normally  runs on cloud servers provided by Microsoft, which can introduce data leak and interception risks. Along those lines, the CIA announced its plan to create a ChatGPT-like service last year, but this Microsoft effort is reportedly a separate project.

Enlarge (credit: Getty Images | Chesnot )

TikTok and its owner ByteDance today sued the federal government to block the "Foreign Adversary Controlled Applications" law that would prohibit TikTok in the US if the company isn't sold to a non-Chinese firm. The complaint in the US Court of Appeals for the District of Columbia Circuit alleges that the law is unconstitutional and asks for a court order prohibiting enforcement.

TikTok and ByteDance say the law "would allow the government to decide that a company may no longer own and publish the innovative and unique speech platform it created. If Congress can do this, it can circumvent the First Amendment by invoking national security and ordering the publisher of any individual newspaper or website to sell to avoid being shut down."

The law will "silenc[e] the 170 million Americans who use the platform to communicate in ways that cannot be replicated elsewhere," TikTok and ByteDance alleged.