Byrne joined Meta in September 2021.

She and her team helped draft the rulebook that applies to the world’s most diabolical people and groups: the Ku Klux Klan, cartels, and of course, terrorists. Meta bans these so-called Dangerous Organizations and Individuals, or DOI, from using its platforms, but further prohibits its billions of users from engaging in “glorification,” “support,” or “representation” of anyone on the list.

Byrne’s job was not only to keep dangerous organizations off Meta properties, but also to prevent their message from spreading across the internet and spilling into the real world. The ambiguity and subjectivity inherent to these terms has made the “DOI” policy a perennial source of over-enforcement and controversy.

A full copy of the secret list obtained by The Intercept in 2021 showed it was disproportionately comprised of Muslim, Arab, and southeast Asian entities, hewing closely to the foreign policy crosshairs of the United States. Much of the list is copied directly from federal blacklists like the Treasury Department’s Specially Designated Global Terrorist roster.

Byrne tried to focus on initiatives and targets that she could feel good about, like efforts to block violent white supremacists from using the company’s VR platform or running Facebook ads. At first she was pleased to see that Meta’s in-house list went further than the federal roster in designating white supremacist organizations like the Klan — or the Azov Battalion.

She was also unsure of whether Meta was up to the task of maintaining a privatized terror roster. “We had this huge problem where we had all of these groups and we didn’t really have … any sort of ongoing check or list of evidence of whether or not these groups were terrorists,” she said, a characterization the company rejected.

Byrne quickly found that the blacklist was flexible. "Meta’s censorship systems are “basically an extension of the government...”

Dec 2 (Reuters) - Intel (INTC.O) Chief Executive Pat Gelsinger has stepped down less than four years after taking the helm of the company, handing control to two lieutenants as the faltering American chipmaking icon searches for a permanent replacement. Gelsinger, who resigned on Dec. 1, left the company before the completion of an ambitious and costly four-year plan to restore the company's lead in making the fastest and smallest computer chips, a crown it lost to Taiwan Semiconductor Manufacturing Co (2330.TW) which makes chips for Intel rivals such as Nvidia (NVDA.O) While Gelsinger has assured both investors and U.S. officials, who are subsidizing Intel's turnaround, that his manufacturing plans remain on track, the full results will not be known until next year, when the company aims to bring a flagship laptop chip back into its own factories.

https://archive.is/aC1ym

https://www.researching.cn/articles/OJ97a9c4dcc97720f3

cross-posted from: https://futurology.today/post/2910566

Alibaba's Qwen team just released QwQ-32B-Preview, a powerful new open-source AI reasoning model that can reason step-by-step through challenging problems and directly competes with OpenAI's o1 series across benchmarks.

The details:

QwQ features a 32K context window, outperforming o1-mini and competing with o1-preview on key math and reasoning benchmarks.

The model was tested across several of the most challenging math and programming benchmarks, showing major advances in deep reasoning.

QwQ demonstrates ‘deep introspection,’ talking through problems step-by-step and questioning and examining its own answers to reason to a solution.

The Qwen team noted several issues in the Preview model, including getting stuck in reasoning loops, struggling with common sense, and language mixing.

Why it matters: Between QwQ and DeepSeek, open-source reasoning models are here — and Chinese firms are absolutely cooking with new models that nearly match the current top closed leaders. Has OpenAI’s moat dried up, or does the AI leader have something special up its sleeve before the end of the year?

cross-posted from: https://lemmy.bestiver.se/post/123708

Comments

Meta is actively helping self-harm content to flourish on Instagram by failing to remove explicit images and encouraging those engaging with such content to befriend one another, according to a damning new study that found its moderation “extremely inadequate”.

Danish researchers created a private self-harm network on the social media platform, including fake profiles of people as young as 13 years old, in which they shared 85 pieces of self-harm-related content gradually increasing in severity, including blood, razor blades and encouragement of self-harm.

The aim of the study was to test Meta’s claim that it had significantly improved its processes for removing harmful content, which it says now uses artificial intelligence (AI). The tech company claims to remove about 99% of harmful content before it is reported.

But Digitalt Ansvar (Digital Accountability), an organisation that promotes responsible digital development, found that in the month-long experiment not a single image was removed.

When it created its own simple AI tool to analyse the content, it was able to automatically identify 38% of the self-harm images and 88% of the most severe. This, the company said, showed that Instagram had access to technology able to address the issue but “has chosen not to implement it effectively”.

cross-posted from: https://lemmy.bestiver.se/post/121053

Comments

Over the past few years, the UEFI threat landscape, particularly that of UEFI bootkits, has evolved significantly. It all started with the first UEFI bootkit proof of concept (PoC) described by Andrea Allievi in 2012, which served as a demonstration of deploying bootkits on modern UEFI-based Windows systems, and was followed with many other PoCs (EfiGuard, Boot Backdoor, UEFI-bootkit). It took several years until the first two real UEFI bootkits were discovered in the wild (ESPecter, 2021 ESET; FinSpy bootkit, 2021 Kaspersky), and it took two more years until the infamous BlackLotus – the first UEFI bootkit capable of bypassing UEFI Secure Boot on up-to-date systems – appeared (2023, ESET).

A common thread among these publicly known bootkits was their exclusive targeting of Windows systems. Today, we unveil our latest discovery: the first UEFI bootkit designed for Linux systems, named Bootkitty by its creators. We believe this bootkit is merely an initial proof of concept, and based on our telemetry, it has not been deployed in the wild. That said, its existence underscores an important message: UEFI bootkits are no longer confined to Windows systems alone.

The bootkit’s main goal is to disable the kernel’s signature verification feature and to preload two as yet unknown ELF binaries via the Linux init process (which is the first process executed by the Linux kernel during system startup). During our analysis, we discovered a possibly related unsigned kernel module – with signs suggesting that it could have been developed by the same author(s) as the bootkit – that deploys an ELF binary responsible for loading yet another kernel module unknown during our analysis.

Supporters of Palestine have called to boycott the payment platform Stripe after its CEO and co-founder Patrick Collison - an Irish-American billionaire who has advocated for Palestinians in the past - posted on social media on Wednesday about his run on the beach in Tel Aviv and how it was "great" to be back.

Many responded to his post on X by pointing out that he was only thirty minutes away from the Gaza Strip. Conservative estimates say nearly 44,000 Palestinians have been killed in Israel's war on Gaza and two million have been under constant Israeli aggression as they fight what UN experts have called "'deliberate starvation".

Some drew comparisons to the Academy Award-winning film, Zone of Interest, which depicts the everyday lives of Germans who lived next to the Auschwitz concentration camp during World War Two.

Hahahahahahahahahahahahaha

https://archive.ph/dBVSG