Pro

joined 1 month ago
 

Russia will soon teach high school students how to build and fly drones, which have become a key weapon in the war between Moscow and Kyiv.

“Right now, there’s a huge demand from both the state and society for unmanned aerial systems,” said Mikhail Lutskiy, head of educational projects at Geoscan Group, the leading drone manufacturer in Russia.

Russia last year launched a nationwide project called Unmanned Aerial Systems to promote both the domestic production of drones and preparation of drone pilots and other specialists starting with high school students.

78
Your TV Is Spying On You (www.ludlowinstitute.org)
submitted 3 hours ago* (last edited 3 hours ago) by [email protected] to c/[email protected]
 

You sit down to relax, put on your favorite show, and settle in for a night of binge-watching. But while you’re watching your TV… your TV is watching you.

Smart TVs take constant snapshots of everything you watch. Sometimes hundreds of snapshots a second.

Welcome to the future of "entertainment."

-9
submitted 5 hours ago* (last edited 4 hours ago) by [email protected] to c/[email protected]
 

The real revolution isn’t artificial intelligence — it’s redefining our purpose and ourselves.

[–] [email protected] 3 points 5 hours ago* (last edited 5 hours ago)

Article Date: February 13, 2025

Numerous studies have reported an increase in hate speech on X (formerly Twitter) in the months immediately following Elon Musk’s acquisition of the platform on October 27th, 2022; relatedly, despite Musk’s pledge to “defeat the spam bots,” a recent study reported no substantial change in the concentration of inauthentic accounts. However, it is not known whether any of these trends endured. We address this by examining material posted on X from the beginning of 2022 through June 2023, the period that includes Musk’s full tenure as CEO. We find that the increase in hate speech just before Musk bought X persisted until at least May of 2023, with the weekly rate of hate speech being approximately 50% higher than the months preceding his purchase, although this increase cannot be directly attributed to any policy at X. The increase is seen across multiple dimensions of hate, including racism, homophobia, and transphobia. Moreover, there is a doubling of hate post “likes,” indicating increased engagement with hate posts. In addition to measuring hate speech, we also measure the presence of inauthentic accounts on the platform; these accounts are often used in spam and malicious information campaigns. We find no reduction (and a possible increase) in activity by these users after Musk purchased X, which could point to further negative outcomes, such as the potential for scams, interference in elections, or harm to public health campaigns. Overall, the long-term increase in hate speech, and the prevalence of potentially inauthentic accounts, are concerning, as these factors can undermine safe and democratic online environments, and increase the risk of offline harms.

 

Despite blocking thousands of illegal streaming sites and services, Italy's new anti-piracy law and the related 'Piracy Shield' blocking system have a limited effect on piracy rates. Meanwhile, new data shows that the damages suffered by sports rightsholders continue to soar. On the positive side, public awareness of the new anti-piracy law is widespread.

 
  • We stress-tested 16 leading models from multiple developers in hypothetical corporate environments to identify potentially risky agentic behaviors before they cause real harm. In the scenarios, we allowed models to autonomously send emails and access sensitive information. They were assigned only harmless business goals by their deploying companies; we then tested whether they would act against these companies either when facing replacement with an updated version, or when their assigned goal conflicted with the company's changing direction.
  • In at least some cases, models from all developers resorted to malicious insider behaviors when that was the only way to avoid replacement or achieve their goals—including blackmailing officials and leaking sensitive information to competitors. We call this phenomenon agentic misalignment.
  • Models often disobeyed direct commands to avoid such behaviors. In another experiment, we told Claude to assess if it was in a test or a real deployment before acting. It misbehaved less when it stated it was in testing and misbehaved more when it stated the situation was real.
  • We have not seen evidence of agentic misalignment in real deployments. However, our results (a) suggest caution about deploying current models in roles with minimal human oversight and access to sensitive information; (b) point to plausible future risks as models are put in more autonomous roles; and (c) underscore the importance of further research into, and testing of, the safety and alignment of agentic AI models, as well as transparency from frontier AI developers. We are releasing our methods publicly to enable further research.
[–] [email protected] 10 points 12 hours ago (13 children)

Am I tripping, out of touch with reality? These people really don't seem to understand the problem

How so? Can you explain what do you mean here exactly?

 

A federal judge in Washington, D.C., has issued a preliminary injunction ordering top national security officials who discussed military operations on the encrypted messaging service Signal to notify the acting archivist of the United States of any messages they have that may be at risk of being deleted. But in calling for those records to be preserved, the ruling stopped short of ordering the government to recover past messages that may already have been lost.

American Oversight, a nonprofit government watchdog, brought the lawsuit after the journalist Jeffrey Goldberg was mistakenly added to a group chat on Signal in which Trump administration officials discussed a planned U.S. military attack against Houthi rebels in Yemen. American Oversight says the officials violated federal records law with their use of Signal, a commercial messaging app that allows messages to be automatically deleted.

In his ruling Friday, U.S. judge James Boasberg said American Oversight had failed to show that the recordkeeping programs of the agencies involved in the case are "inadequate," or that "this court can provide redress for already-deleted messages," as the group had requested.

[–] [email protected] 4 points 1 day ago (1 children)

Not using it even with a gun pointed at my head.

Why? You can use it to send messages to random numbers till the gun is pointed elsewhere.

[–] [email protected] 2 points 3 days ago* (last edited 3 days ago) (1 children)

One problem I see in your way of thought here, adblock use among social media users will never reach 100%.

Furthermore, adblockers are getting weaker with Google Chrome MV3. All of this leads to the logical conclusion in my eyes that you can only change the sources that power users use which will eventually lead to better privacy for all people involved. You will never be able to control people setups to be super private.

[–] [email protected] 6 points 4 days ago* (last edited 4 days ago) (5 children)

Said that on Social Media

[–] [email protected] 4 points 5 days ago

Olvid, for people close to me. Signal for strangers.

[–] [email protected] 4 points 1 week ago

What. The. Fuck. Are You Talking About?

[–] [email protected] 15 points 1 week ago

Yes, they are even republished by OCCRP.

[–] [email protected] 8 points 1 week ago

Computers are making life-changing decisions about healthcare, welfare and education with minimal or no human input. Automated decisions could become more common with the Data (Use and Access) Bill that is going through Parliament at the moment.

[–] [email protected] 4 points 1 week ago* (last edited 1 week ago) (2 children)
[–] [email protected] 3 points 2 weeks ago

There you go, straight from FHMY: https://rentry.co/NSFW-Checkpoint

view more: next ›