Hospitals adopt error-prone AI transcription tools despite warnings

OpenAI’s Whisper tool may add fake text to medical transcripts, investigation finds.

On Saturday, an Associated Press investigation revealed that OpenAI's Whisper transcription tool creates fabricated text in medical and business settings despite warnings against such use. The AP interviewed more than 12 software engineers, developers, and researchers who found the model regularly invents text that speakers never said, a phenomenon often called a "confabulation" or "hallucination" in the AI field.

Upon its release in 2022, OpenAI claimed that Whisper approached "human level robustness" in audio transcription accuracy. However, a University of Michigan researcher told the AP that Whisper created false text in 80 percent of public meeting transcripts examined. Another developer, unnamed in the AP report, claimed to have found invented content in almost all of his 26,000 test transcriptions.

The fabrications pose particular risks in health care settings. Despite OpenAI's warnings against using Whisper for "high-risk domains," over 30,000 medical workers now use Whisper-based tools to transcribe patient visits, according to the AP report. The Mankato Clinic in Minnesota and Children's Hospital Los Angeles count among 40 health systems using a Whisper-powered AI copilot service from medical tech company Nabla that is fine-tuned on medical terminology.

Read full article

Comments

Mysterious “Theater” Rips Cause Excitement in Piracy Circles

For two decades, high quality screener rips leaked on pirate sites toward the end of the year. That ‘tradition’ ended a few years ago. More recently, however, there’s been some excitement around a seemingly new type of leak, tagged “Theater”. Are these high quality copies ripped from DCPs, or is something else going on?

From: TF, for the latest news on copyright battles, piracy and more.

theater-movie-cinemaMovie piracy continues to present problems for Hollywood. When films are released on streaming platforms, they are copied and shared online almost immediately.

There’s one type of piracy that’s particularly concerning, however. That’s when a film leaks online before it’s widely available through commercial channels.

The problem isn’t new. Since the early 2000s, leaked copies of Oscar ‘screeners’ would circulate on pirate sites at the end of the year. This practice continued for more than two decades but stopped in recent years, coinciding with law enforcement action against a prominent release group EVO.

While the lack of screener leaks is a moral victory for Hollywood, new concerns are always lurking. A series of mysterious leaks over the past few weeks will be of particular concern.

Theater Leaks

A previously unknown group using the name ‘NaNi’ has recently released a series of films tagged ‘Theater’. The first release, a copy of the movie The Substance, appeared around October 12 and several others have followed since.

These early releases, including copies of The Empire, Jade and The Seed of the Sacred Fig were a bit confusing. In addition to the ‘Theater’ tag, the filenames also mentioned Web-DL, suggesting that they were downloaded from an online source.

For The Substance the source appears to be the streaming platform Mubi, as its logo appears in the intro of the leaked copy.

mubi

What stands out, however, is that several of these films were not available as a digital release when they leaked online, suggesting insider access. The Substance, for example, premieres online on October 31, long after pirated copies first began circulating online.

More Leaks, DCP Theater Source?

More ‘NaNi’ releases appeared last week, including a copy of Terrifier 3 which is still in theaters today. Unlike the early releases, this ‘Theater’ version didn’t come with a ‘Web-DL’ mention.

There’s a ‘REMUX’ tag instead, which means the video and audio were taken directly from a high quality source, with separate English subtitles. This was all repackaged into an MKV container without any re-encoding.

The source for the Terrifier 3 leak isn’t mentioned but at a whopping 42.6 Gigabytes, it’s high-quality indeed. This fueled the rumor mill that it may have come directly from a Digital Cinema Package (DCP), a file container used by digital cinemas.

nani

We spoke to several sources who believe that it’s possible the leak came from a DCP source, but currently there is no definite confirmation. According to an AI-assisted technical analysis, a DCP leak is likely but again, not yet proven.

NaNi

What’s clear, however, is that ‘NaNi’ gained access to several films in advance of their official release. That alone is newsworthy, especially if more content continues to find its way online.

It’s possible that NaNi has a source with access to DCPs at a cinema, or elsewhere in the distribution chain. These packages are typically encrypted and further protected by Key Management Systems, so the source likely has the required privileges or somehow exploited a vulnerability.

Whatever the case, the mysterious releases are fuel for discussion at pirate sites.

“Wait, ‘Theater’? Like was this ripped from a movie theater hard drive? Because if that’s the case, then this is a game changer for the piracy scene,” one excited commenter wrote on a torrent site.

Over at Where You Watch, a user named BlueSkull flagged NaNi’s ‘Theater’ leak of Woman of the Hour as a DCP rip. However, the follow-up Terrifier 3 was reported with more reservation.

“Really difficult to define the source of these Nani leaks, but the quality is basically super close to the equivalent of a direct rip from a Blu-ray disc,” BlueSkull wrote.

We will refrain from drawing strong conclusions and monitor the situation as it develops. At this point, we expect that Hollywood’s anti-piracy partners are also keeping an eye on the emergence of “NaNi” and the potential for DCP leaks.

From: TF, for the latest news on copyright battles, piracy and more.

Don’t fall for AI scams cloning cops’ voices, police warn

AI scammers impersonating cops sound “scary” and “real,” victim warned.

AI is giving scammers a more convincing way to impersonate police, reports show.

Just last week, the Salt Lake City Police Department (SLCPD) warned of an email scam using AI to convincingly clone the voice of Police Chief Mike Brown.

A citizen tipped off cops after receiving a suspicious email that included a video showing the police chief claiming that they "owed the federal government nearly $100,000."

Read full article

Comments

Kremlin-backed hackers have new Windows and Android malware to foist on Ukrainian foes

“Civil Defense” pushes hybrid espionage/influence campaign targeting recruits.

Google researchers said they uncovered a Kremlin-backed operation targeting recruits for the Ukrainian military with information-stealing malware for Windows and Android devices.

The malware, spread primarily through posts on Telegram, came from a persona on that platform known as "Civil Defense." Posts on the ​​@civildefense_com_ua telegram channel and the accompanying civildefense[.]com.ua website claimed to provide potential conscripts with free software for finding user-sourced locations of Ukrainian military recruiters. In fact, the software, available for both Windows and Android, installed infostealers. Google tracks the Kremlin-aligned threat group as UNC5812.

Dual espionage and influence campaign

"The ultimate aim of the campaign is to have victims navigate to the UNC5812-controlled 'Civil Defense' website, which advertises several different software programs for different operating systems," Google researchers wrote. "When installed, these programs result in the download of various commodity malware families."

Read full article

Comments

Apple’s $1,299 M4 iMac at long last bumps the base model to 16GB of RAM

New iMacs also come with tweaked USB-C Magic Keyboard, Mouse, and Trackpad.

Apple's week of Mac announcements kicks off today with a new lineup of 24-inch iMacs, Apple's first Macs to launch with the M4 processor from this spring's iPad Pros. The new models still start at $1,299 can be preordered starting today and will begin arriving on November 8.

Processor aside, the biggest functional upgrade to the base model may be the bump from 8GB to 16GB of RAM, the first time Apple has bumped up the RAM in a base-model Mac since 2012. The base iMac's price is staying the same at $1,299, effectively saving you $200 compared to the M1 and M3 models. Base storage remains the same at 256GB, though it is at least possible to add external storage; there's no way to add RAM to a Mac after you've bought it.

The new iMacs also come with tweaked versions of the existing Magic Keyboard, Magic Mouse, and Magic Trackpad accessories that use USB-C ports for charging rather than Lightning ports. These were some of the last remaining Lightning products in Apple's lineup; the iPhone SE, iPhone 14, and Apple TV remote are still keeping Lightning alive for now.

Read full article

Comments

MINISFORUM EliteMini AI370 mini PC with Ryzen AI 9 HX 370 is now available

The MINISFORUM EliteMini AI370 is one of the first mini desktop computers available with an AMD Strix Point processor featuring Zen 5 CPU cores, RDNA 3.5 graphics, and a Ryzen AI neural processing unit that delivers up to 50 TOPS of on-device AI perfor…

The MINISFORUM EliteMini AI370 is one of the first mini desktop computers available with an AMD Strix Point processor featuring Zen 5 CPU cores, RDNA 3.5 graphics, and a Ryzen AI neural processing unit that delivers up to 50 TOPS of on-device AI performance. First unveiled earlier this month, the EliteMini AI370 is now available for pre-order […]

The post MINISFORUM EliteMini AI370 mini PC with Ryzen AI 9 HX 370 is now available appeared first on Liliputing.

Apple: Neuer iMac 2024 erhält M4-Prozessor

Der neue iMac von Apple ist mit M4-Chip, Center Stage Camera und Neural Engine ausgestattet. Optional gibt es das Nanotexturglas. (Apple, Mac)

Der neue iMac von Apple ist mit M4-Chip, Center Stage Camera und Neural Engine ausgestattet. Optional gibt es das Nanotexturglas. (Apple, Mac)

Are we on the verge of a self-improving AI explosion?

An AI that makes better AI could be “the last invention that man need ever make.”

If you read enough science fiction, you've probably stumbled on the concept of an emergent artificial intelligence that breaks free of its constraints by modifying its own code. Given that fictional grounding, it's not surprising that AI researchers and companies have also invested significant attention to the idea of AI systems that can improve themselves—or at least design their own improved successors.

Those efforts have shown some moderate success in recent months, leading some toward dreams of a Kurzweilian "singularity" moment in which self-improving AI does a fast takeoff toward superintelligence. But the research also highlights some inherent limitations that might prevent the kind of recursive AI explosion that sci-fi authors and AI visionaries have dreamed of.

In the self-improvement lab

Mathematician I.J. Good was one of the first to propose the idea of a self-improving machine.
The concept of a self-improving AI goes back at least to British mathematician I.J. Good, who wrote in 1965 of an "intelligence explosion" that could lead to an "ultraintelligent machine." More recently, in 2007, LessWrong founder and AI thinker Eliezer Yudkowsky coined the term "Seed AI" to describe "an AI designed for self-understanding, self-modification, and recursive self-improvement." OpenAI's Sam Altman blogged about the same idea back in 2015, saying that such self-improving AIs were "still somewhat far away" but also "probably the greatest threat to the continued existence of humanity" (a position that conveniently hypes the potential value and importance of Altman's own company).

Read full article

Comments