According to their studies, the older we get, the more we will match our name. Wild, but interesting theory.

cross-posted from: https://midwest.social/post/15413291

Perceptive announced it performed the world's first fully automated dental procedure using a robot. The system integrates AI, 3D imaging, and robotics, enabling tasks like crown placements in just 15 minutes. Perceptive claims their robotic dentist can diagnose and plan treatments using a 3D volumetric data procedure and an optical coherence technology (OCT) scan. Perceptive's CEO, Chris Ciriello, mentioned that FDA approval could take up to five years.

I don't usually self promo but I was interested in the concept of a social archipelago facilitated by closed federation/allowlisting, in response to a lot of the bigotry and spam that's on the Fediverse and how difficult it can be to moderate. I was also curious about how Beehaw/Lemmy users feel about allowlisting and closed federation especially since Beehaw's on the cusp of switching to an allowlist. What are y'all's thoughts on the concept of a social archipelago?

While writing, I consulted these two essays which introduce this idea, so I'll leave them here as "further reading":

• The Fediverse is Already Dead
• Islands: An Opt-In Federated Network

cross-posted from: https://feddit.org/post/1626469

The European Commission has made TikTok's commitments to permanently withdraw TikTok Lite Rewards programme from the EU binding. These commitments have been submitted by TikTok to address the concerns raised by the Commission in the formal proceedings opened against TikTok on 22 April and ensure compliance with the Digital Services Act (DSA).

“The available brain time of young Europeans is not a currency for social media — and it never will be,” said Thierry Breton, Commissioner for Internal Market.

This is the first case the Commission closes under the DSA, 105 days after the opening of the proceedings.

Maybe they'll take each other out.

Hed o' the day so far ...

Archive link

Nvidia scraped videos from Youtube and several other sources to compile training data for its AI products, internal Slack chats, emails, and documents obtained by 404 Media show.

When asked about legal and ethical aspects of using copyrighted content to train an AI model, Nvidia defended its practice as being “in full compliance with the letter and the spirit of copyright law.” Internal conversations at Nvidia viewed by 404 Media show when employees working on the project raised questions about potential legal issues surrounding the use of datasets compiled by academics for research purposes and YouTube videos, managers told them they had clearance to use that content from the highest levels of the company.

Emails from the project’s leadership to employees show that the goal of Cosmos (different from the company’s existing Cosmos deep learning product) was to build a state-of-the-art video foundation model “that encapsulates simulation of light transport, physics, and intelligence in one place to unlock various downstream applications critical to NVIDIA.”

Slack messages from inside a channel the company set up for the project show employees using an open-source YouTube video downloader called yt-dlp, combined with virtual machines that refresh IP addresses to avoid being blocked by YouTube. According to the messages, they were attempting to download full-length videos from a variety of sources including Netflix, but were focused on YouTube videos. Emails viewed by 404 Media show project managers discussing using 20 to 30 virtual machines in Amazon Web Services to download 80 years-worth of videos per day.

The original article is gated (requires free registering)

Autocracies and weak democracies import Chinese surveillance AI during times of unrest in pursuit of greater political control, a study by the Prohect Syndicate shows.

Since China has aggressively deployed AI-powered facial recognition to support its own surveillance state, the repirt sets out to explore the patterns and political consequences of trade in these technologies. After constructing a database for global trade in facial recognition AI from 2008 to 2021, the study comprises 1,636 deals from 36 exporting countries to 136 importing countries.

• Autocracies and weak democracies are more likely to import facial recognition AI from China. While the US predominantly exports the technology to mature democracies (these account for roughly two-thirds of its links, or three-quarters of its deals), China exports roughly equal amounts to mature democracies and autocracies or weak democracies

• When comparing China’s exports of facial-recognition AI to its exports of other frontier technologies, we found that facial recognition AI is the only technology for which China displays an autocratic bias. Equally notable, we found no such bias when investigating the US.

• One potential explanation for this difference is that autocracies and weak democracies might be turning specifically to China for surveillance technologies. Autocracies and weak democracies are more likely to import facial recognition AI from China in years when they experience domestic unrest.

• Imports of Chinese surveillance AI during episodes of domestic unrest are indeed associated with a country’s elections becoming less fair, less peaceful and less credible overall. A similar pattern appears to hold with imports of US surveillance AI, although this finding is less precisely estimated.

• This suggests a need for tighter AI trade regulation, which could be modeled on the regulation of other goods that produce negative externalities. Insofar as autocratically-biased AI is trained on data collected for the purpose of political repression, it is similar to goods produced from unethically-sourced inputs, such as child labor. Since surveillance AI might have negative downstream externalities, such as lost civil liberties and political rights, it is not unlike pollution.

Similar to all dual-use technologies, facial recognition AI has the potential to benefit consumers and firms. However, regulations must be carefully designed to ensure that this frontier technology is diffused around the world without facilitating autocratization.

Archived version

Within hours of a local 17-year-old boy being arrested for the mass-stabbings in Southport, a seaside town in northwest England, untrue narratives started circulating on social media naming him as “Ali al-Shakati”—a Muslim migrant to the UK—alleging that he was on an MI6 watchlist, and that he was an asylum seeker who was known to the Liverpool mental health services.

None of this was true, but research by Dr Marc Owen Jones, an expert in digital authoritarianism, has traced how this kind of speculation rapidly notched up 27m impressions on social media.

The self-proclaimed misogynist and alleged rapist Andrew Tate, with nearly 10m followers on X, posted a false image of the supposed attacker, claiming he was “straight off a boat”—even though by then the police had told us he had been born in Cardiff 17 years ago. But that, according to Tate, was a lie promoted by what he calls “the Matrix”.

One of the most prominent amplifiers of this untrue information was a shadowy organisation calling itself Channel3 Now. Quite who is behind this outfit is unclear. Investigative journalists soon found that it had started life as a place for Russian car rally videos. It may be now run out of an address in Pakistan or the US. That’s the joy of Musk’s beloved “independent media”—you haven’t got a clue who half of the fabulists are.

[...]

At the end of July 2024, the destructive #it'snotmyukraine was among the three most popular hashtags on the TikTok Creative Centre website, the Ukrainian media outlet Texty has found.

The numbers were impressive: 41,000 posts in a week and 10 million views — incredible results for a hashtag that didn't exist a month ago.

"We know for sure that hundreds of identical profiles on TikTok have been created with one goal in mind - to distribute videos about the misconduct of Recruiting office or actors playing Recruiting office representatives to the widest possible audience," the organization says. "We are not even considering the veracity of the videos themselves, as we are talking about 2,100 accounts that are trying to bypass the platform's algorithms and go viral among Ukrainians."

Fake accounts carefully try to redirect users to the "Renaissance of Ukraine" telegram channel. Created on 7 July, the channel posts "up-to-date information about Recruiting office, their actions and lawlessness. Useful tips and tricks", as well as calls for information to be sent to the bot and promises to create an up-to-date map with Recruiting office employees in different cities of Ukraine.

Sometimes, it also compares the 2014 Olympic Games in Paris and Russia. With the classic propaganda question, "Why is the West promoting this gayness and debauchery?"

An interesting little back and forth between Codeberg and someone advocating for them to spend resources on AI assistants.

The last post in the thread, for easier consumption: https://mastodon.social/@s_levi_s/112903065829205736

I need to change ISPs and need to find a new email provider. This time I want to move to my own domain which I purchased through Namecheap and I do not want to use another ISP's email system nor do I want to use Google, or Microsoft since I am Linux (and Android too) based. I would like this to be US based or at least have a strong US presence so obvious choices like Proton Mail, Mailfence, and Mailbox.org are out. I would prefer it interoperate well with FOSS software too, I use Thunderbird and K-9 Mail for example. Also so want them to be trustworthy, have good security, and have good OpSec with respect to their their servers and service.

After looking I find three I am considering and they are quite different:

• Fastmail. Long history. No PGP support but they do have their own domains one can use also.
• Namecheap Private Email. Uses Ox App Suite, may support PGP, and quite new. I think you have to have your own domain (not sure).
• Forward Email (forwardemail.net). A forwarder with IMAP support. You supply the webmail if you want webmail, but otherwise it should work fine with IMAP and normal clients.

So questions:

• Any thoughts and experience, pros and cons with the above 3.
• Other better ideas.

So thoughts? Thanks.

Feedback welcome! Here's the TL;DR list

1. Listen more to more Black people
2. Post less – and think before you post
3. Call in, call out, and/or report anti-Blackness when you see it
4. Support Black people and Black-led instances and projects

Other suggestions?

researchers conducted experimental surveys with more than 1,000 adults in the U.S. to evaluate the relationship between AI disclosure and consumer behavior

The findings consistently showed products described as using artificial intelligence were less popular

“When AI is mentioned, it tends to lower emotional trust, which in turn decreases purchase intentions,”

Archived version

The U.S. Department of Justice is suing TikTok for allegedly letting kids under 13 make accounts without their parents’ permission and collecting “extensive data” on them, in violation of US child privacy law.

The DOJ claims that TikTok knowingly let kids onto its platform through its “Kids Mode,” collected their information, and failed to delete their accounts at their parents’ requests, in violation of the Children’s Online Privacy Protection Act (COPPA). When a kid under 13 entered their age on the app, they’d be prompted to enter a username, which doesn’t contain personal information, and it would create a Kids Mode account for the user. But the app wouldn’t notify parents or get their consent. Kids can’t upload videos in that mode, but they can view videos; the DOJ alleges that TikTok collected some personal information on them as part of this process, like unique device identifiers and IP addresses.

The lawsuit alleges that TikTok’s age-gating techniques “are deficient in multiple ways.” Under an earlier practice, TikTok would let users restart the account creation process even if they’d originally entered a birthday showing they’re under 13, according to the complaint. TikTok also used to let users log in through Instagram or Google, which would categorize the accounts as “age unknown,” the DOJ alleges.

Archived link

Research into websites that are openly advertising services to a cybercriminal audience, such as bulletproof hosting, reveals that many of these domains are supported by Cloudflare’s services, the NGO Spamhaus says.

For years, Spamhaus has observed abusive activity facilitated by Cloudflare’s various services. Cybercriminals have been exploiting these legitimate services to mask activities and enhance their malicious operations, a tactic referred to as living off trusted services (LOTS).

With 1201 unresolved Spamhaus Blocklist (SBL) listings, it is clear that the state of affairs at Cloudflare’s Connectivity Cloud looks less than optimal from an abuse-handling perspective, Spamhaus writes on its website. 10.05% of all domains listed on Spamhaus’s Domain Blocklist (DBL), which indicates signs of spam or malicious activity, are on Cloudflare nameservers . Spamhaus routinely observes miscreants moving their domains, which are already listed in the DBL, to Cloudflare to disguise the backend of their operation, be it spamvertized domains, phishing, or worse.