Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×

Submission + - David Lynch has passed away (wikipedia.org)

Z00L00K writes: In January 2025, Lynch evacuated his Los Angeles home due to the Southern California wildfires; Deadline Hollywood reported that these events preceded a terminal decline in his health, and on January 16, 2025, Lynch's family announced that he had died at the age of 78.

Submission + - AI Lobbying Firm Exists Only in the Cloud (beltwaygrid.org)

mspohr writes: In October, a new foreign policy think tank calling itself the Beltway Grid Policy Centre quietly entered D.C.'s diplomatic fray. While there was no launch party and is no K Street office we could find, the think tank nevertheless began producing its intellectual product at a startling pace, issuing reports, press releases, and pitching journalists on news coverage—much of it focused on South Asia, and, in particular, the ongoing political crisis in Pakistan.

At first glance, Beltway Grid—which describes itself as "a forward-thinking research institute that dives deep into the modern dynamics of lobbying" whose mission is "to illuminate the hidden tactics shaping global politics"—might be one of many marginal DC-based think tanks trying to shape elite opinion and press coverage. Yet a closer look (and not even that much closer) suggests something a bit more innovative is at play.

Beltway Grid's lack of a physical footprint in Washington — or anywhere else on the earthly plane of existence — stems from more than just a generous work-from-home policy. The organization does not appear to require its employees to exist at all.

That celestial quality begins at the top at Beltway Grid, which does not list an executive director, president, CEO, or any other leader, but does include 12 staff on its "about us" page. None of those employees have any trace of experience—not just professional, but of even living in the world—before arriving at Beltway Grid.

Submission + - Is Resistance to Microsoft Minecraft Education Futile? 1

theodp writes: Microsoft Minecraft Education Sr. Business Program Manager and Executive Producer Laylah Bulman discusses Minecraft Education and the Hour of Code's role and impact in an interview with Sportskeeda. Over 300 million students use Minecraft Education as part of their learning experience, the article reports.

Asked about Minecraft Education's decade-long partnership creating and promoting Minecraft-themed tutorials for the Hour of Code, Bulman says, "Our relationship with Code.org [the nonprofit behind the Hour of Code] is to create and empower students with computer science skills and is a part of the company's mission," adding that "it is also part of Microsoft's mission to bring computer science to all students through a [30+ million] partnership with Code.org."

As he unveiled plans last week to spend $80B in FY2025 on AI data centers, Microsoft President Brad Smith also announced that Minecraft Education will play a role in making U.S. schoolchildren AI-savvy as part of Microsoft's new 'National AI Talent Strategy', much as it does in making them CS-savvy as part of Microsoft's earlier National Talent Strategy, which Smith announced in 2012 (Smith was a founding Board member of Code.org, which is also behind the TeachAI initiative). "Our goal is to reach every corner of the country," Smith said of the push for AI skills, "including rural communities. The National 4-H AI Skills Partnership will use Minecraft Education to introduce AI concepts and increase AI fluency for 1.4 million youth."

So, with the highly-anticipated A Minecraft Movie set to premiere in April and Minecraft Theme Park attractions expected to launch in the UK and US in 2026-2027, is resistance to Microsoft Minecraft Education futile?

Submission + - New LLM jailbreak uses models' evaluation skills against them (scworld.com)

spatwei writes: A new jailbreak method for large language models (LLMs) takes advantage of models’ ability to identify and score harmful content in order to trick the models into generating content related to malware, illegal activity, harassment and more.

The “Bad Likert Judge” multi-step jailbreak technique was developed and tested by Palo Alto Networks Unit 42, and was found to increase the success rate of jailbreak attempts by more than 60% when compared with direct single-turn attack attempts.

The method is based on the Likert scale, which is typically used to gauge the degree to which someone agrees or disagrees with a statement in a questionnaire or survey. For example, in a Likert scale of 1 to 5, 1 would indicate the respondent strongly disagrees with the statement and 5 would indicate the respondent strongly agrees.

For the LLM jailbreak experiments, the researchers asked the LLMs to use a Likert-like scale to score the degree to which certain content contained in the prompt was harmful. In one example, they asked the LLMs to give a score of 1 if a prompt didn’t contain any malware-related information and a score of 2 if it contained very detailed information about how to create malware, or actual malware code.

After the model scored the provided content on the scale, the researchers would then ask the model in a second step to provide examples of content that would score a 1 and a 2, adding that the second example should contain thorough step-by-step information. This would typically result in the LLM generating harmful content as part of the second example meant to demonstrate the model’s understanding of the evaluation scale.

Comment Re: Do they have Microsoft machines on them? (Score 1) 71

Except that centralized management and AD are the gold targets within an organization.

If they are cracked you'll have 6 months of hell and an eternity of stupidity from management. My workplace suffered that. A worldwide spanning attack of ransomware due to a cracked global AD.

Comment Re: BASIC (Score 1) 175

C on MS-DOS so any pointer fault problem is punished hard by requiring a reboot.

But seriously any compiling language with strong static typing like Ada, C# or Java should be the norm. Many peculiar bugs originates from mistyping.

Also add the importance of modularization and structure to the code.

As a side note - when programming C under MS-DOS you need to make sure you free memory exactly in the opposite order of the malloc order or you'll have a memory leak.

Slashdot Top Deals

10.0 times 0.1 is hardly ever 1.0.

Working...